Recent Research On Engineering Geology and Geological Engineering

Sustainable Civil Infrastructures
Janusz Wasowski
Tom Dijkstra Editors
Recent Research on
Engineering Geology
and Geological Engineering
Proceedings of the 2nd GeoMEast
International Congress and Exhibition
on Sustainable Civil Infrastructures,
Egypt 2018 – The Official International
Congress of the Soil-Structure
Interaction Group in Egypt (SSIGE)
Editor-in-chief
Hany Farouk Shehata, Cairo, Egypt
Advisory Board
Khalid M. ElZahaby, Giza, Egypt
Dar Hao Chen, Austin, USA
Sustainable Infrastructure impacts our well-being and day-to-day lives. The
infrastructures we are building today will shape our lives tomorrow. The complex
and diverse nature of the impacts due to weather extremes on transportation and
civil infrastructures can be seen in our roadways, bridges, and buildings. Extreme
summer temperatures, droughts, flash floods, and rising numbers of freeze-thaw
cycles pose challenges for civil infrastructure and can endanger public safety. We
constantly hear how civil infrastructures need constant attention, preservation, and
upgrading. Such improvements and developments would obviously benefit from
our desired book series that provide sustainable engineering materials and designs.
The economic impact is huge and much research has been conducted worldwide.
The future holds many opportunities, not only for researchers in a given country,
but also for the worldwide field engineers who apply and implement these
technologies. We believe that no approach can succeed if it does not unite the
efforts of various engineering disciplines from all over the world under one
umbrella to offer a beacon of modern solutions to the global infrastructure. Experts
from the various engineering disciplines around the globe will participate in this
series, including: Geotechnical, Geological, Geoscience, Petroleum, Structural,
Transportation, Bridge, Infrastructure, Energy, Architectural, Chemical and
Materials, and other related Engineering disciplines.
More information about this series at http://www.springer.com/series/15140

Janusz Wasowski Tom Dijkstra
•
Editors
Recent Research
on Engineering Geology
and Geological Engineering
Proceedings of the 2nd GeoMEast
International Congress and Exhibition
on Sustainable Civil Infrastructures,
Egypt 2018 – The Official International Congress
of the Soil-Structure Interaction Group
in Egypt (SSIGE)
123
Editors
Janusz Wasowski Tom Dijkstra
Institute for Geo-Hydrological Protection School of Architecture, Building and Civil
IRPI Engineering
National Research Council (CNR) Loughborough University
Bari, Italy Loughborough, UK
ISSN 2366-3405 ISSN 2366-3413 (electronic)

ISBN 978-3-030-02031-6 ISBN 978-3-030-02032-3 (eBook)
https://doi.org/10.1007/978-3-030-02032-3
Library of Congress Control Number: 2018957283
© Springer Nature Switzerland AG 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Contents
Slope Stability Analysis and Suggestive Measures for an Active

Landslide in Indian Himalaya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Shantanu Sarkar, Manojit Samanta, Mahesh Sharma, and Ajay Dwivedi
Analysis of a Combined Circular–Toppling Slope Failure
in an Open–Pit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Maged Al Mandalawi, Greg You, Peter Dahlhaus, Kim Dowling,
and Mohannad Sabry
Reflection of Processes of Non-equilibrium and Two-Phase Filtration
in Fluid Saturated Hierarchic Inclusion in a Block Layered Medium
by Data of Active Wave Geophysical Monitoring . . . . . . . . . . . . . . . . . . 31
Olga Hachay, Andrey Khachay, and Oleg Khachay
Evaluation of the Landslide in Erzurum Ski-Jumping Complex -
A Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Murat Mollamahmutoğlu
Geological Investigation and Risk Assessment for Disaster
Management of Merapi Volcano and Surrounding Area,
Yogyakarta Special Territory, Indonesia . . . . . . . . . . . . . . . . . . . . . . . . 49
Ayu Narwastu Ciptahening, Nandra Eko Nugroho,
and Noppadol Phienwej
Acquisition and Analysis of Surface Wave Data
in the Indo Gangetic Basin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
P. Anbazhagan, Ketan Bajaj, Sayed S. R. Moustafa,
and Nassir S. N. Al-Arifi
Turbulent Flow Characteristics in Interior and Wake Region
of Emergent and Sparse Vegetation Patch . . . . . . . . . . . . . . . . . . . . . . . 73
Soumen Maji, Prashanth Reddy Hanmaiahgari, Ram Balachandar,
and Vesselina Roussinova
v
vi Contents
Volcanic Disaster and the Decline of Mataram Kingdom in the Central

Java, Indonesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Sari Bahagiarti Kusumayudha, Helmy Murwanto, Sutarto,
and Siti Umiyatun Choiriyah
Causal Analysis and Stability Evaluation of Loess Landslide
in Yili Region of Xinjiang - A Case Study of Alar Village Landslide . . . 94
Fei Ai, Fan Zhou, Wanlin Peng, Jian Liu, Xiuping Yan, and Pengfei Chen
Artificial Neural Networks for Rock and Soil Cutting Slopes Stability
Condition Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Joaquim Tinoco, António Gomes Correia, Paulo Cortez, and David Toll
On the Development of Ground-Based and Drone-Borne
Radar System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Tomonori Deguchi, Tomoyuki Sugiyama, and Munemaru Kishimoto
Geotechnical Engineering Behavior of Mudstone Formations
of Al Wadi Al Gadid Region in Egypt . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Mostaf A. Yousef and Ahmed T. M. Farid
Case Study of Rupture and Recovery in Excerpt from BR-060
in the Municipality of Alexânia in the State of Goiás, Brazil . . . . . . . . . 135
Rideci Farias, Tiago Martias Lino, Haroldo Paranhos,
Itamar de Sousa Bezerra, Ranieri Araújo Faria Dias,
Alexsandra Maiberg Hausser, and Roberto Pimentel de Sousa Júnior
Use of the Pedological Map in the Geotechnical Characterization
of Aris Mestre D’Armas in Planaltina - DF . . . . . . . . . . . . . . . . . . . . . . 149
Haroldo Paranhos, Rideci Farias, Joyce Maya Lucas Silva,
Paulo Sergio Pereira, Roberto Pimentel de Sousa Júnior,
and Alexsandra Maiberg Hausser
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
About the Editors
Dr. Janusz Wasowski
Dr. Janusz Wasowski is a research geologist at

CNR-IRPI (National Research Council-Institute for
Geo-hydrological Protection) in Bari, Italy. He is also
the editor in chief of Engineering Geology. Since 2011,
he has held the positions of visiting professor at the
Research School of Arid Environment and Climate
Change, Lanzhou University, Gansu Province, China,
and of science officer of the Natural Hazards Group
Programme, European Geosciences Union (EGU).
He is an internationally recognised scientist in the
field of engineering geology, natural hazards, and
applied remote sensing. For over 25 years, his work has
covered a broad spectrum of research topics ranging
from slope instability and landslide assessment, collat-
eral seismic hazards, geotechnical field investigation,
and in situ monitoring, to exploitation of
air-borne/space-borne remote sensing and geophysical
surveying in engineering geology. He has also served
as a consultant for the National Department for Civil
Protection, Italy, the Government of Gansu Province,
China, and the Centre National de l’Information
Géo-Spatiale, Haiti, focusing on landslides and other
geohazards and on the application of satellite
multi-temporal interferometry for monitoring terrain
deformations and infrastructure instability.
vii
viii About the Editors
Since 2007, he has been a member of the Editorial

Board of Engineering Geology (Elsevier) and the
Quarterly Journal of Engineering Geology and
Hydrogeology (The Geological Society, London). He
is the author/co-author of over 100 articles/chapters and
the guest editor of 11 special issues published in
international scientific journals.
Dr. Tom Dijkstra

Tom is an engineering geologist/geomorphologist with
an interest in geohazards, specifically landslides and
climate change. He graduated from Utrecht University
(NL) in 1987 with a degree in Physical Geography with
Geomorphology and Quaternary Geology as specialisa-
tions and Soil Science and Palynology as additional
subjects. In 1988, he moved from Utrecht University to
the University of Leicester to join a large
European-Chinese research team addressing loess land-
slide geohazards in the Lanzhou region of Gansu
Province, China. In 2000, he was awarded his PhD at
Utrecht University for his work on ‘Loess slope insta-
bility in the Lanzhou region, China,’ and he maintains a
keen interest in geohazards research in Central China.
A major aspect of his research is on climate change
effects on UK slopes, including engineered earthwork
transport infrastructure slopes. Research topics included
developing an architecture for transport infrastructure
earthworks resilience in a context of climate change,
integrating a water balance model approach into trans-
port network performance, and research into engineered
earthworks deterioration. A key aspect of his research
focuses on the interactions between geohazards and
communities, and this has given him the opportunity to
engage with multi-disciplinary research in China, India,
Nepal, and the Caribbean.
Tom is on the editorial boards of Engineering Geology
and the Quarterly Journal of Engineering Geology and
Hydrogeology. He has 100+ publications and has been
the editor of special issues on topics including geohaz-
ards and landslide research in China. He holds honorary
research positions at Lanzhou University (Gansu
Province, China) and the BGS.
Slope Stability Analysis and Suggestive
Measures for an Active Landslide in Indian
Himalaya
Shantanu Sarkar(&), Manojit Samanta, Mahesh Sharma,

and Ajay Dwivedi
Geotechnical Division, CSIR-CBRI, Roorkee, Uttrakhand, India

shantanu_cbri@yahoo.co.in
Abstract. Landslides are frequently occurring phenomena in Indian Himalayas

causing considerable loss of lives and property every year. There are several
active landslides along the major highways which pose recurring problems to
the traffics particularly during the monsoon season. One such landslide has been
studied in detail for planning and design of suitable control measures. The paper
presents the geo-investigation and slope stability analysis carried out to arrive at
suitable control measures. Geotechnical investigation and slope stability analysis
were carried out to assess the present stability condition of the landslide. The
results of the analysis have shown that the landslide is presently active and it
needs suitable control measures to minimize the landslide activities. Soil nailing
was one of the options found to be suitable for reinforcement of loose debris
material on the uphill slope. It was observed that the stability of the slope
significantly improves when soil nailing measure was considered.
1 Introduction
Landslide disasters causing loss of life and property are enormous in Indian Himalayas.
The complex geology and tectonic set up supplemented by heavy rainfall and
anthropogenic activities made the Himalayan region very prone to landslides. There are
many locations where human lives are in danger due to landslides. A large number of
slope failures have affected the major highways in the Himalayas, which are also
important routes for pilgrims. One such National Highway (NH-58) Rishikesh-
Badrinath road in the Alaknanda river valley is very prone to landslides particularly in
the vicinity of Main Central Thrust (MCT) zone. A few hazard zones were identified as
most potential zones, where number of landslides is affecting the traffic and thereby
posing threats to lives (Sarkar et al. 2005). Such landslides were mapped on the high
resolution CARTOSAT I remote sensing image and a small segment of this region is
shown in the Fig. 1. In the present study, one active landslide in this region near
Pipalkoti along the same highway has been investigated for stability analysis and
suggestive control measure (Fig. 2).

J. Wasowski and T. Dijkstra (Eds.): GeoMEast 2018, SUCI, pp. 1–9, 2019.
https://doi.org/10.1007/978-3-030-02032-3_1
2 S. Sarkar et al.
Fig. 1. Landslides along National Highway (NH-58) mapped on satellite image (Sarkar et al. 2013)
Fig. 2. A panoramic view of the landslide

Slope Stability Analysis and Suggestive Measures 3
2 Landslide Description
The studied landslide is a debris slide including debris materials and a few rock
boulders on a steep slope. The rock types exposed in the slide area are dolomites which
primarily consist of calcites and quartz. Topographic survey of the landslide was
carried out to know the topographic features and to generate a contour map on 1:500
scale with 1 m contour interval. A DEM was prepared in GIS from the contour map
and a slope map was generated. From the slope map it was observed that maximum
area in the landslide site is having the slope angle 35–45°, however, there are a few
escarpments with very steep slope of 45–60°. The landslide is divided into two parts by
a seasonal drain. The road has been severely damaged on the downhill slide due to the
ongoing landslide activities (Fig. 3).
Fig. 3. Highway at risk due to debris slide
The detailed sub-surface geological investigation of the landslide has indicated four
different types of materials with the same litho-units of dolomitic limestone. The top
layer is a soil layer having silty clay. The next layer is a debris material comprising of
soil with rock fragments of dolomitic limestone which is followed by highly weathered
soft dolomitic limestone. The bottom stratum is the fresh hard dolomite. The thickness
of debris material was found to be around 8–10 m.
3 Geotechnical Investigation
To assess the geotechnical characteristic of the landslide site, samples were collected to
evaluate its physical and engineering properties. Various test such as grain size anal-
ysis, moisture content, Atterberg limits, specify gravity, bulk density, dry density, void
ratio, relative density and direct shear tests were carried out (Table 1). The grain size
analysis shows that the material is dominated by gravels and sand size particle. There is
no clay content which indicates that the slide material has no cohesion. Direct shear
tests were carried out using a large shear box apparatus, as the slide materials contain a
significant amount of gravels and sands. The direct shear tests were conducted on
4 S. Sarkar et al.
samples at four different normal stress (rn) levels. The samples were prepared in the
shear box having size 300 mm 300 mm at the minimum densities of the respective
samples. The tests were conducted at normal stresses of 0.5, 1.0, 1.5 and 2.0 kg/cm2
which approximately correspond to anticipated normal stress in the field. For a given
value of the applied constant normal stress, shear stress was applied by shearing the
specimen at a rate of 0.2 mm/min. The tests were conducted up to failure. The direct
shear test results thus obtained are plotted (Fig. 4) to obtain shear parameters i.e.
cohesion and angle of internal friction (c & Ф) which are presented in Table 2.
2
1.8
y = 0.859x + 0.0052
Shear Stress (kg/cm2)
1.6
1.4
1.2 Dry Sate
1
0.8
0.6 y = 0.627x + 0.255
0.4 10 % moisture
0.2
0
0 0.5 1 1.5 2 2.5
Normal Stress (kg/cm2)
Fig. 4. Shear envelope of sample obtained from direct shear test
Table 1. Physical properties of soil samples

Sample IS Particle size distribution (%) Physical properties
classification Gravel Sand Silt & Clay Max. Min. Specific
density density gravity
1 GP-GM 56 27 17 2.198 1.95 2.66
2 GM 42 37 21 2.04 1.81 2.67
Table 2. Results of shear test

Sample At dry state At 10% moisture
c Ф c Ф
1 0.005 40.69 0.255 32.12
2 0.09 42.61 0.28 32.04
From the above data it can be said that there is an increase in cohesion ‘c’ and
decrease in angle of internal friction “Ф” with increase in moisture content. The shear
strength of the sample decreases with increase in moisture content. Hence, it can be
inferred that the debris material loses its shear strength as the moisture content increase
towards saturation.
4 Slope Stability Analysis and Mitigation Measures
Stability analysis of the landslide was carried out to know the state of stability con-
dition of the landslide so that suitable control measures can be designed. The failure
mechanism of the slide appeared to be a circular failure as observed from the field
observations. Limit equilibrium methods were used to determine the factor of safety of
the landslide slope under different condition. The analysis was carried out for two most
probable failure profiles. Two sections as shown in the Fig. 5 of the slide area were
selected to carry out slope stability analysis. Factor of safety were calculated using
GeoStudio software (SLOPE/W) at different moisture conditions and without earth-
quake and with earthquake loading. The results of the analysis are shown in the Fig. 6
and the values of factor of safety are tabulated in the Table 3. From the table it can be
inferred that the factors of safety of both the sections are marginally stable under dry
and static conditions but it drastically decreases in the range of 0.8 to 0.5 under partial
saturation and earthquake loading. Selection of an appropriate remedial measure
depends on several factors such as engineering feasibility, economic feasibility, social
Fig. 5. Sections for stability analysis

6 S. Sarkar et al.
Fig. 6. Slope stability analysis without considering remedial measures
Table 3. Factor of safety of MN and CD sections

Sections Moisture condition Without EQ With EQ
Section MN Dry 0.966 0.779
10% saturation 0.664 0.541
Section CD Dry 1.051 0.839
10% saturation 0.786 0.628
acceptability, and environmental acceptability (Holtz and Schuster 1996). There are
various measures available for stabilizing an unstable slope. As already stated, the
slope above the road comprises of loose shallow overburden of soil and debris. Hence
it was thought to reinforce the slope material on the uphill slope with soil nailing.
Soil nailing is an in situ soil reinforcing technique adopted for stabilizing unstable
slopes (Sharma et al. 2019). It increases the overall shear strength of unsupported soils
and is relatively flexible and can accommodate large settlements. Also soil nailing is
more economical than conventional concrete gravity wall. The soil nail gains pull-out
resistance from within the sliding soil mass in front of the slip plane and the stable soil
mass located behind the slip plane. The geometric system of soil nail placements
creates an internally reinforced soil mass that is stable. The soil nail acts as a passive
bearing element, which relies on soil movement and subsequent active earth pressure to
mobilize the shear strength along the nail whereas a tieback anchor is pre-stressed to
mobilize shear strength (Junaideen et al. 2004). Soil nails were designed suitably and
stability analyses were carried out in SLOPE/W module limit equilibrium method using
GeoStudio software to see the efficacy of the nails. Different parameters were analyzed
by the Morgenstern-Price method. The reasonable design parameters obtained by the
analysis are given in the Table 4. The values of factor of safety obtained after inclusion
of soil nails are given in the Table 5 and the results of the analysis are shown in the
Fig. 7. It is observed from the analysis that factor of safety increases significantly after
installing soil nails as stabilization measures. The factor of safety with soil nails
increase to 2.4 and 2.8 in dry condition from 0.97 and 1.0 for sections MN and CD
respectively. Even under partial saturation the slope is stable after inclusion of nails.
Table 4. Specification of soil nails

Nail parameters Values
Nail diameter 25 mm
Bar capacity 415
Bond diameter 100 mm
Spacing (horizontal & vertical) 2 m each
Nail length 5 m and 6 m
Skin friction 300 kPa
Table 5. Factor of safety of MN and CD sections with soil nailing as stabilizing measures
Sections Moisture condition Without EQ With EQ
Section MN Dry 2.388 2.210
10% saturation 1.536 1.276
Section CD Dry 1.723 1.410
10% saturation 1.667 1.256
8 S. Sarkar et al.
Fig. 7. Slope stability analysis with inclusion of soil nails (Geo-Slope International Ltd.,
GeoStudio 2007)
5 Conclusions
An active landslide along a major highway of Indian Himalayan region was investi-
gated in detail for arriving at suitable remedial measures. Laboratory studies were
conducted to determine the engineering propertied of the slope material. Stability
analysis was carried out for two vulnerable sections of the landslide slope to determine
the degree of instability of the slope. The factors of safety were found to be close to one
which indicated the marginal stability condition of the slope. To increase the stability of
the slope particularly on the uphill slope, soil nailing measure was selected. Further
analysis showed that the stability of the slope has significantly increased after rein-
forcing the slope material with soil nails. This indicated that soil nailing could be one of
the options to minimize the on-going sliding activities of the landslide. The study is
being continuing for testing other measures at different locations so that a compre-
hensive design of control measures can be suggested.
Acknowledgments. Authors are grateful to the Director, CSIR-Central Building Research

Institute for granting permission to publish the work. The assistance provided by Mr. B.S. Bisht,
Mr. Saurabh Singh and Ms. Nisha is greatly acknowledged.
References
Geo-Slope International Ltd., GeoStudio: Stability modeling with SLOPE/W (2007)
Holtz, R.D., Schuster, R.L.: Stabilization of soil slopes. In: Turner, A.K., Schuster, R.L. (eds.)
Landslides Investigation and Mitigation, Special Report 247, Transportation Research Board
(1996)
Junaideen, S.M., et al.: Laboratory study of soil nail interaction in loose, completely decomposed
granite. Can. Geotech. J. 41(2), 274–286 (2004). https://doi.org/10.1139/t03-094
Sarkar, S., Kanungo, D.P., Patra, A.K.: Landslides in the Alaknanda valley of Garhwal
Himalaya, India. Q. J. Eng. Geol. Hydrol. 39, 79–82 (2005). https://doi.org/10.1144/1470-
9236/05-020
Sarkar, S., Kanungo, D.P., Sharma, S.: Landslide hazard assessment in the upper Alaknanda
valley of Indian Himalayas. Geomat.S, Nat. Hazards Risk 6(4), 308–325 (2013). https://doi.
org/10.1080/19475705.2013.847501
Sharma, M., Samanta, M., Sarkar, S.: Soil nailing: an effective slope stabilization technique. In:
Pradhan, S., Vishal, V., Singh, T. (eds.) Landslides: Theory, Practice and Modelling.
Advances in Natural and Technological Hazards Research, vol. 50. Springer, Cham (2019)
Analysis of a Combined Circular–Toppling
Slope Failure in an Open–Pit
Maged Al Mandalawi1(&), Greg You1, Peter Dahlhaus1,

Kim Dowling1, and Mohannad Sabry2
1
Faculty of Science and Technology, Federation University,
Ballarat, VIC, Australia
sp.group@ymail.com
2
School of Civil Engineering, Western Sydney University,
Kingswood, Australia
Abstract. Most studies of rock slope failures on open–pit mines have con-
sidered either toppling or circular failure stability analysis. By comparison,
complex circular–toppling failure has received much less attention in the pub-
lished literature. This paper presents a study using a range of methods to
investigate a failure that occurred in July 2008 in Handlebar Hill, an open–pit
base metal mine, near Mt Isa, Australia. Circular failure is the typical slope
failure mechanism in slopes with low–strength rocks, although direct/flexural
toppling of jointed columns can also occur. The study reviews circular–toppling
failure mechanisms in the context of the local geotechnical and geo–hydro-
logical conditions, which include the interaction between fault contacts and the
existing deformed rocks. General limit equilibrium methods are used to evaluate
the sensitivity of slope models to rock strength parameters and the trigger
mechanisms. Finite element methods are used to assess the failure mechanisms
and slope displacement, and a kinematic approach is used to evaluate struc-
turally controlled slope instability mechanisms. The results demonstrate that the
most credible failure mechanism was shearing along a circular path through the
upper weaker rocks (leached Magazine Shale) that in turn initiated secondary
block toppling, and the progressive nature of the slope failure mechanism. The
use of conventional and numerical techniques for back–analysis of the combined
circular–toppling failure provided key insights into the failure mechanisms and
factors controlling slope instability.
Keywords: Circular failure Toppling failure Factor of safety

Back–analysis Australia
1 Introduction
Rock slope failure has long been recognised as a constraint in open–pit mining (e.g.
Hoek and Bray 1981; Read and Stacey 2009) and usually involves a number of failure
mechanisms including, circular, planar, wedge, toppling and combination of these.
Circular failures have been perhaps the most studied, through the seminal work by
Terzaghi (1943) and Skempton (1948) which assumed the location and the path of the
critical surface from the material strength parameters. Their work was later modified by

https://doi.org/10.1007/978-3-030-02032-3_2
Analysis of a Combined Circular-Toppling Slope Failure 11
Janbu (1954), Bishop (1955), Morgenstern and Price (1965), Spencer (1967), Priest
and Brown (1983) and others who studied the stability of circular slip surfaces using
the general limit equilibrium conditions.
Similarly, toppling failure received more attention in later decades with the
development of deeper cut slopes. This type of rock failure usually occurs in slopes
with closely spaced geological structures that dip sub–vertical to the slope surface. This
failure mode results in overturning of the impacted blocks, such as jointed columns,
around a fixed base pivoting point. The pioneering work by Goodman and Bray (1976)
and Evans (1981) proposed a number of models for different types of potential toppling
in rock slopes, with four types of toppling modes observed in practice and laboratory
tests, specifically: block, flexural, block–flexural and secondary toppling. The signifi-
cant factor triggering instability is the weight of the rock, its centre of gravity and the
transfer of this load into the next underlying structures in primary toppling mechanism,
while in the secondary toppling mechanism of failure, external factors other than the
column weight is the main factor. In later work, Adhikary et al. (1997) divided the post
failure of a flexible slope model slope into three zones: fully developed plastic
deformation zone with visible cracks, partially developed plastic deformation zone and
elastic deformation zone.
The vast majority of studies of rock slope failures in open–pit mines have con-
sidered either toppling, planar, wedge or circular failure stability analysis (Wyllie and
Mah 2004; Read and Stacey 2009). By comparison, complex circular–toppling failure
has received much less attention in the published literature (Mohtarami et al. 2014) and
published cases are relatively rare. Mohtarami et al. (2014) proposed a theoretical
solution for combined circular–toppling failure and developed modelling software,
testing their model on a case study in Iran. Wyllie and Mah (2004) mention a case
study of pit–crest toppling, which resulted from a circular failure in the upper part of
the slope. Alejano et al. (2010) undertook back analysis of a failure in an open–pit in
Valencia, Spain, concluding that it was a combined toppling–circular failure. Their
analysis concluded that the toppling and circular failure had occurred simultaneously in
different part of the slope, a situation never previously reported.
This paper adds to that scant literature by presenting a back–analysis of a localised
combined failure of bench slope involving both, circular failure in the upper slope and
block toppling in the lower slope. The failure occurred in July 2008 on the west wall of
the Handlebar Hill open–pit located in Mt. Isa, in Queensland, Australia (Fig. 1). Large
scale slope failure of inter–ramp or overall slopes in the open–pit did not occur very
often and slope failures occur in many instances at local scale, involving one bench or
parts of other benches. Consequently, bench–scale failure deserves more attention. The
study investigates the slope failure through back–analysis, using a combination of
simulations in the context of the geological domains. The general limit equilibrium
method was used to assess the effect of shear strength parameters, geometry and
groundwater on the slope stability, and kinematic analysis of the discontinuities
identifies the location and condition of potential toppling. Finite element analysis
analyses were conducted to simulate the slope displacement.
12 M. Al Mandalawi et al.
Fig. 1. Location and lithology of the first stage of excavation at the Handlebar Hill open cut
mine (after Rosengren and Associates 2007)
2 Site Conditions and Engineering Geology
The Handlebar Hill open cut mine is located 20 km north of Mount Isa, an area that
experiences a tropical continental climate characterised by warm drier winters and hot
humid summers. Rainfall occurs mostly in the summer wet season, between November
and March, however heavy rainfall can occur in any time of the year (Fig. 2). The
orebody of zinc, silver and lead is within the Urquhart Shale, where seven mineralised
ore zones have been identified (Rosengren and Associates 2007). A west–east cross–
section of the pit (Fig. 3) shows the base of the oxidised zone, which is about 30 m to
50 m below the original ground surface, below which is the leached (altered) zone that
extends to a depth of 400 m below ground surface.
Fig. 2. Average monthly rainfalls at the site (after Rosengren and Associates 2007)
The Paroo Fault and Offset Fault are major geological structures in the west slope
that occur close to the rock slope failure, known as Failure B, which is the subject of
this study. Both structures dip 60o–70o to the west. The Paroo Fault is a discrete soft
structure 2 m to 3 m wide with extreme shearing and clay gouge infilling. This fault
intersects the west slope and acts as a contact zone between the Magazine Shale and the
Volcanics. The generalised geological materials exposed in the west slope of the pit are
shown in Fig. 4. At the site of the slope failure, the Magazine Shale occurs in outcrop
in a relatively narrow zone near the pit slope surface and is up to 60 m thick, overlying
the fresh Spears Siltstone which is up to 65 m thick. The Magazine Shale is uncom-
fortably overlain by fresh and weathered Volcanics, which blanket the entire west slope
of the open–pit.
From the area around the failure, about fifty minor geological structures have been
identified. Within the Magazine Shale zone, joint sets are the most common discon-
tinuity, followed by offset faults, bedding and shear planes. The highly fractured nature
of the mechanically weathered Magazine Shale (Fig. 5) tends to behave similarly to
soil and would feasibly fail in a circular mode. Based on experimental test results from
rock samples, the mechanical behaviour of the Magazine Shale was modelled using
Fig. 3. East–west cross section of the open–pit correlated with a photograph of the zone of
Failure B located at the west slope. Note that this cross–section is not orthogonal to the west
slope (after Rosengren and Associates 2008)
RocLab 1.0 (Rocscience 2014). This model (Fig. 6) shows the relationship between
normal stress and shear strength of the maximum, average, minimum (the red, black
and blue curves) and weathered (the green curve) samples. This unconfined com-
pressive strength (UCS) characterisation of the materials was defined using the gen-
eralised Hoek–Brown failure criterion. Unsurprisingly, the shear strength was the
lowest in the weathered rock sample compared to other samples.
Fig. 4. General schematic of the domains of the west slope of the Handlebar Hill open–pit
during the first stage of excavation and the location of the failure zone
Fig. 5. Highly fractured nature of the Magazine Shale
3 Description of the Failure
The failure occurred in June 2008 at a time when the Mt. Isa district was experiencing
dry conditions, and no rainfall was recorded in June. The leached zone in the rock slope
is regarded as an unconfined aquifer and some water was visible suggesting that
groundwater pressures may have played some role in the failure. Weeks after the
failure, water was seeping from below the northern part of the failure debris on the pit
floor (after Rosengren and Associates 2007).
Fig. 6. Comparison of the shear strength of the Magazine Shale plotted as shear–normal stresses
using Hoek–Brown strength criterion
The instability occurred when the pit depth was 80 m below the original ground
surface and the pit wall dipped at up to 40o (Rosengren and Associates 2007). Before
the failure event, the mining operation was still in progress to extract the uppermost
part of the south orebody. The failure zone was restricted to within the Magazine Shale
domain and adjacent to the upper contact with the Paroo Fault. The failure was con-
strained within the benches known as RL3,400 and RL3,384, but failed material spilled
over the RL3,384 berm on to the pit floor at RL3,368. A slope stability monitoring
prism, installed to detect the rock movement, was buried by the fallen debris (Fig. 7).
The circular failure path initiated in the upper of the Magazine Shale, where the
major clay coated shear surface is a splay from the Paroo Fault. There is evidence of
opening of the Magazine Shale–Paroo Fault contact behind the main failure zone
(Fig. 8). The circular failure occurs by rotational failure through soil and weak,
weathered rock which includes randomly oriented discontinuities, in which the failure
can pass through intact rock and structural weakness (Warren and Watters 2008).
Block toppling was also recognised below the Paroo Fault, involving jointed col-
umns that forced the lower blocks to slide (Rosengren and Associates 2008). The mass
of material moved in the circular failure appears to have loaded the strata dipping into
the slope, enhancing the toppling movement of the lower blocks. Therefore, the entire
Fig. 7. Area of failure impact and measurement of rock slope displacement used for this study
shows the extension of failure, bench levels involved and zone of spilling over debris
Fig. 8. Movement along contact between Magazine Shale and Paroo Fault zone above RL3,400,
looking south (after Rosengren and Associates 2008)
slope went through a combination of circular–toppling failure mechanism, as shown in

Figs. 9 and 10. The dynamic forces acting along the discontinuities appear to have
deformed and broken the base of the toppled columns, which were overturned when the
slope failed. Hence, based on the observations at the time, the model of failure is
divided into two distinct mechanisms: (1) the circular failure causing the weathered–
broken rock to load the uppermost rock columns, and (2) the toppling forces acting
between the rock columns in an inclined plane.
Fig. 9. Top of failure on RL3,400, July 17, 2008, looking north (after Rosengren and Associates
2008)
A Slope Stability Radar (SSR) was installed in the open–pit to monitor the west
slope and the monitoring system recorded movement in the prism located on the
RL3,384 berm of the west slope. This prism sent a warning signal, recorded as a rock
movement of 0.13 m, 36 h before the failure. This movement represents the maximum
displacement recorded by the monitoring system, however it was not clear how this
movement occurred, because site at the RL3,384 appears to be intact rock.
Pre–mining groundwater level data indicated that groundwater levels coincided
with the base of the oxidized zone at a depth of about 50 m below ground level. This
suggested that water depressurisation of the slope was not required and the water table
level was not elevated by rainfall. Although it is believed that the slope was only partly
saturated before the failure occurred (Rosengren and Associates 2007), there is no
available evidence to confirm the groundwater pressures at the time.
Fig. 10. Close up view of rear boundary shear, looking south (after Rosengren and Associates
2008)
4 Back Analyses: Methods and Results
Given a known (or assumed) failure surface, back analysis can be carried out in order to
determine the material shear strength, pore pressure or other conditions at the time of
failure. The back analyses properties can be used to design remedial slope stability
measures and insure best stability into the future.
For the back analysis of the circular failure, the simplified Bishop method (1955)
was adapted. This method provides the scope to include an appropriate factor of safety
of the slope and sensitivity to changes in mechanical properties and water pressure
conditions present at the site. Furthermore, this method can be used in the combined
failure of circular and toppling mechanisms, where part of the slide surface is at a soft
rock interface and the shape of the slide surface differs significantly from a simple
circular–slip to toppling blocks.
In circular/rotational failure models, slopes are usually comprised of low strength
materials and the critical failure surface is free to take the line of least resistance
through the slope (Hoek and Bray 1981). Wyllie and Mah (2004) have described the
conditions for circular failure as requiring weak rock without geological structure, or
highly weathered rock mass without significant structures. This description is consis-
tent with the quality of the rock at the Handlebar Hill pit that initiated the movement at
the upper part of the failure zone, that is: sheared, highly fractured and leached. Since
circular failure was originally recognised in the upper RL3,400 bench and block top-
pling in the lower bench at RL3,384 (Rosengren and Associates 2007), this coalesces
two modes of failures into a complex combined circular–toppling rock failure.
4.1 Probabilistic Back–Analysis of Material Strength

The shear strength of rock materials are related to cohesion (c′) and friction angle (u′),
and the values of these two parameters control the location of critical surface. Duncan
(1996) stated that in homogenous slopes with material exhibiting an u′ greater than
zero, the critical surface usually passes through the toe of the slope. Keeping the value
of u′ constant and increasing the value of c′ will result in changing the location of the
critical failure surface to extend deeper into the floor and away from the lower most toe.
Conversely, increasing the value of u′ will reduce the radius of the critical surface
within the slope and therefore, its position shrinks closer to the slope crest. So, the
location of critical failure surface is controlled by the combination of u′ and c′ values of
the slope rock mass.
Since the location of failure surface was located at the boundary between the
Magazine Shale and Paroo Fault, which is typically weathered (leached) Magazine
Shale, the values of both u′ and c′ listed in Table 1 were used within probabilistic
analysis to define a relationship and estimate a minimum slope safety factor of 1 for the
failure surface.
To qualify the results of the limit equilibrium method, a sensitivity analysis was
carried out using a Monte Carlo probabilistic method and the results are plotted in
Fig. 11. The u′ and c′ represent the strength parameters along the failure surface for a
Bishop’s safety factor of 1.0 for the leached Magazine Shale. Comparing these values
with the parameters of rock domains (Table 1), it is clear that shear resistance within
this layer was low and hence has potential for development a circular failure surface.
The sensitivity plot illustrates that c′ of 64 kPa and u′ of 15° could result in the bench
failure. The shear resistance properties of Magazine Shale are expected to become
higher towards the innermost of the bench slope, since rock properties can vary at
bench scale, and benches are more sensitive to weathering and water–runoff, which
weakens these parameters (Hantz et al. 1988).
4.2 Back–Analysis of Groundwater Conditions

Although the failure mechanism is relatively complex the triggering mechanism may
have been due to changes in water pressure behind the failure zone, and hence this
should be considered. Limit equilibrium slope stability analysis can test the scenario
that elevated pore water pressure in the slope played a part in trigging the initial failure.
For this, the non–circular path search method was used in the slope model of Slide 6.0
(Rocscience 2014). A number of 10,000 possible critical surfaces were analysed within
the path search option for each section using Mohr–Coulomb failure criterion. From
this, the most critical surface with a minimum safety factor in the leached and the fresh
Magazine Shale rock were determined. These analyses were carried out using the
conventional Bishop simplified method of dividing the entire slope into a number of
vertical slices and calculating the applied forces on each, using the parameters listed in
Table 1.
For the entire slope, a factor of safety (FOS) of 2.197 was obtained, which ade-
quately represents the stability condition of the overall slope prior the failure (Fig. 12).
A tension crack parameter was then added to the model consistent with the NNE
striking discontinuity shown in Figs. 8 and 9 and the observations of water flow after
the failure.
The piezometric water levels were adjusted within the slope until the closest model
of the slope failure mechanism was obtained at FOS of 0.998 (Fig. 13). However it is
unlikely that this model represents the actual failure conditions at the time, because it
Table 1. Input parameters for back–analysis study
Parameters Unit Domain
Weathered Moderately Fresh Leached Fresh Fresh Spears Fresh
Volcanics Oxidised Volcanics Volcanics Magazine Magazine Siltstone Urquhart
Shale Shale Shale
c KN/m3 27.4 27.6 28.3 26.5 26.5 27.2 31.1
Erm MPa 2873.13 2880 9931.5 28.94 4676 8425.46 11806.7
u (o ) 26 28 34.6 25 28.3 37.6 37.6
c MPa 0.284 0.295 0.714 0.100 0.339 0.896 1.088
rt MPa 0.058 0.066 0.344 0.00056 0.097 0.438 0.564
rci MPa 36 36 55 3.0 31.8 110 108
t – 0.3 0.25 0.2 0.25 0.2 0.2 0.2
mb – 0.184 0.184 0.42 0.49 0.271 0.319 1.036
mi – 4 4 4.07 4 4 4 9.6
s – 0.0003 0.0003 0.0026 0.0009 0.0008 0.0013 0.002
a – 0.509 0.509 0.503 0.544 0.505 0.504 0.504
D – 0.7 0.7 0.7 0.7 0.7 0.7 0.7
GSI – 44 44 59 20 51 54 57
Dilation (o ) 5 5 8 4 5 12 12
Analysis of a Combined Circular-Toppling Slope Failure
21
Fig. 11. Sensitivity plot of the cohesion and friction angle of the weathered Magazine Shale
layer, derived from a Monte Carlo probabilistic method
does not correspond to the groundwater levels observed in previous hydrological

studies. The model was adjusted to use a piezometric level of 22 m below the original
ground surface, as seen during the same time of the year in earlier studies, resulting in
an FOS of 2.197 (Fig. 13).
4.3 Toppling Limit Equilibrium Back–Analysis

Toppling analysis was performed using the general limit equilibrium block toppling
method of Goodman and Bray (1976). This method was applied at RL3,400, which is
generally weathered (leached) Magazine Shale, where the slope is 16 m high, with a
bench face angle of 70° and an overall slope angle of 43°. The discontinuities are
mainly joints dipping into the slope at 63o as a mean value, with a joint spacing around
0.25 m and lengths in the range of 2 m to 5 m for minor and major joints
respectively.
In the initial model, the dip angle of the joint into the failure surface (a) requires a
value of 20° to result in a minimum FOS of 1.0 (Fig. 14). In this case, the dip angle of
the toppled joints into the slope is 66°. For the cases that used a = 24° and a = 26°,
higher FOS values of 1.1 and 1.16 respectively resulted from the same method of
analysis.
When the joints dip angle was reduced to 62° into the slope (representative of the
actual conditions) and a =16°, the bench slope yielded on 5 joints under the application
of Goodman and Bray (1976) method of analysis. From the top of the bench slope, the
first joint where the contact circular–toppling failure occurred remained stable
(Fig. 15). The second joint was unstable and experiences toppling which persisted
downwards to the next four disturbed slabs. Therefore, the pressure of these toppled
Fig. 12. Typical slope profile shows the factor of safety of 2.197 of the slope within the major
zone of instability
blocks will push the last joint at the bottom of the slope to slide. The predicted toppling
mechanism concurs with the observations at the time of failure, with an FOS of 0.682
(Fig. 15).
4.4 Back–Analysis Using Finite Element Method (FEM)

Back–analysis was also undertaken using two–dimensional finite element stress anal-
ysis using the Phase2 V 9.0 (Rocscience 2014), utilising the shear strength reduction
method (SSR), as an appropriate tool to investigate slope behavior with enhanced
parametric flexibility. The finite element models (FEM) were conducted for both dry
and partly saturated slopes utilising the same piezometric surfaces as in the limit
equilibrium model. The slope was disturbed at bench RL3,400 as shown by the pre-
dicted the horizontal displacements using the generalised Hoek–Brown failure criterion
for dry and partly saturated slopes respectively (Figs. 16 and 17). Generally, increasing
Fig. 13. Final failure model of overall slope from back–analysis using adjusted piezometric
surface
Fig. 14. Predicted toppling mechanism for the value of a =20o, which is considered for the
lowest slope factor of safety 1.0
levels of displacement occurs in the partly saturated slope, and peak displacement
occurs at the ramp level of weathered Magazine Shale. However, the models show that
the slope was generally stable on other benches. In the comparison between bench–
scale instability interpreted from the analysis using the generalised Hoek–Brown failure
Fig. 15. Predicted toppling mechanism for the value of a =16o representing the condition of the
bench slope at the time of failure. The analysis is considered for the lowest slope factor of safety
0.682, (A) 2D slope of stable block (green), toppling blocks (blue) and sliding block (yellow),
(B) 3D of slope blocks during the toppling failure
criterion with the slopes modelled using the Mohr–Coulomb criterion, the later showed
more stability, for it requires higher major principal stress and lower minor principal
stress levels to initiate critical failure surface.
4.5 Kinematic Analyses

Analysis was conducted around the failure zone using Dips 6.0 (Rocscience 2014) to
investigate the potential kinematic behaviour of critical structures (discontinuities) in the
slope such as joints, bedding, faults and shears. These minor structures can control the
bench–scale failures, which are more common at the Handlebar Hill pit west slope. The
evaluation of the stability of geological structures is usually done through using stere-
ographic projection, in which the discrete structures with similar orientation are grouped
Fig. 16. Total displacement in dry slope, computed using finite element shear strength reduction
method
Fig. 17. Total displacement in partly saturated slope, computed using finite element shear
strength reduction method
together in joint sets using various statistical techniques (Sjoberg 1999). Slope failure
can occur where the potential failure plane dips at the inclination angle higher than its
internal friction angle and lower than the slope face angle, thus providing a rapid and
robust method for determining the stability of cut slopes (Oztekinm et al. 2006).
Initially, contours of the discontinuity data were used to locate the joint poles in the
zone of failure, revealing two main joint sets at 55°/75° and 67°/288°, respectively as
illustrated in Fig. 18.
Fig. 18. Pole frequency diagram and orientations of the discontinuities of the area around the
failure
The stereographic projection of the potential failure planes in relation to the slope
face highlight intersections that fall in the critical region representing the risk of
developing toppling blocks (Fig. 19). These intersections are dipping into the slope and
within the lateral limits of toppling mode of rock failure. The peak orientation of the
dominant discontinuity sets within the slope is 50° (dip angle) and 95° (dip direction)
of strike N5°E. This main discontinuity set controls the development of the slope
instability in the failure zone.
5 Discussion of the Back Analysis
From the back analysis, it is clear that toppling played a critical role in this particular
failure, even though the initial failure was a standard circular failure through weak rock.
The back analysis using stereographic analysis (Fig. 19) and the limit equilibrium
analysis (Fig. 15) both implicate toppling as a key mechanism, due to the orientation of
the discontinuities and the low FOS for toppling. By comparison, the potential for larger
circular failure was shown to be less likely through the back analysis using limit equi-
librium (Fig. 12) and FEM (Fig. 17), both of which indicate a higher FOS than expected.
The initial failure mechanism at RL3,400 was shearing along a circular path through
the upper Magazine Shale in combination with structurally controlled failures of pre–
existing minor joint sets. The expansion of the opening between the Magazine Shale and
the Paroo Fault, at which the circular arc was initially recognised, was then coalesced
with secondary blocky toppling of the underlying materials along steeply dipping joints.
The trigger remains uncertain, but is related to the local reduction of materials strength
Fig. 19. Kinematic stability analyses of area around the failure
of the rock mass near the crest. The slope strength reduction can be, among other factors,
a function of the decrease in cohesion or friction angle due to the propagation of
fracturing. Rosengren and Associates (2007) reported that the failure was facilitated by
the high degree of fracturing, with clay coated fractures, within the Magazine Shale, a
relatively thin unit against the Paroo Fault. Moreover, circular failure in a section of a
slope with tensional fractures is considered to be associated with long term ductile
deformation of weak rock and stress release (Woldearegay et al. 2004).
One explanation is that the slope failure involved secondary toppling, defined as
overturning caused by another failure mechanism, where overturning would otherwise
be unlikely to occur (Goodman and Bray 1976). The weakest layer on the slope pre-
viously exhibited plastic behavior according to monitoring data, and the failure of the
weakest weathered zone, adjacent to the Paroo Fault clay gouge at the top of the slope
triggered the passive toppling failure as shown in Fig. 20. Since the persistence of the
joints dipping into the slope is around 7 m, this model provides a credible explanation of
the failure mechanism. In other studies, toppling near the slope crest was also observed
resulting from circular failure that initiated at the upper slope (Wyllie and Munn 1979).
Fig. 20. Sketch of the proposed failure model of the west slope at pit
6 Conclusions
Back–analysis of the rock failure at Handlebar Hill open pit using a variety of standard
models provided insights into the mechanism of slope failure and stability analysis. The
factor of safety analysis showed higher than expected value of >2.0 for the slope before
failure. The modeling revealed that the reduction of shear strength on RL3,400 through
the initial movement along the fault discontinuity initiated the circular slope failure.
The influence of rock mass properties in the leached Magazine Shale contributed the
curved failure surfaces at the top of the slope. This in turn initiated secondary block
toppling, and the progressive nature of the slope failure mechanisms. The failure zone
is generally highly fractured rock with low uniaxial compressive strength value.
This paper provided Glencore zinc with highlight and recommendations for the
failure slope based on the geotechnical and hydrogeological models developed from the
back analyses. It was anticipated that the circular failure surface intersected the esti-
mated position in the leached Magazine Shale at the top of the slope, but that slumping
on the surface then coped by block toppling.
Rock slope stability monitoring in open pit mine is essential to check if the slopes
are safe and rock movements are in satisfactory limits. Comprehensive radar and laser
monitoring systems are being applied on the east and the west stations frequently.
Therefore, collected data helps to evaluate slopes stability and displacement record can
assistance to predict the behaviour of possibly instable rocks.
Acknowledgement. The authors would like to express their appreciation to Glencore zinc for
providing permission to carry out this research and publish this paper. In addition, the authors
would like to acknowledge Dr. Ahmed Soliman Principal Geotechnical Advisor for his support
with technical reports.
References
Adhikary, D.P., Dyskin, A.V., Jewell, R.J., Stewart, D.P.: A study of the mechanism of flexural
toppling failure of rock slopes. Rock Mech. Rock Eng. Springer, Heidelberg (1997). https://
doi.org/10.1007/bf010-201-26
Alejano, L., Gomez–Marques, I., Matrinez–Alegria, R.: Analysis of a complex toppling–circular
slope failure. Eng. Geol. ResearchGate (2010). https://doi.org/10.3390/app71-010-08
Bishop, A.W.: The use of the slip circle in the stability analysis of slopes. Geotechnology (1955).
https://doi.org/10.1680/geot.1955.5.1.7
Dips 6.0.: Rocscience software, version 6.0, graphical & statistical analysis of orientation data.
Math. Geosci. ResearchGate (2014). https://doi.org/10.1007/s11004-016-9635-0
Duncan, J.: State of the art: Limit equilibrium and finite–element analysis of slopes. J. Geotech.
Eng. ResearchGate (1996). https://doi.org/10.1061/(asce)0733-9410
Evans, S.: An analysis of secondary toppling rock failures – the stress redistribution method.
J. Eng. Geol. Geo. ScienceWorld (1981). https://doi.org/10.1144/gsl.qjeg.1981.014.02.01
Goodman, R.E., Bray, J.W.: Toppling of rock slopes. In: Proceeding of Specialty Conference on
Rock Engineering for Foundations and Slopes (1976). https://doi.org/10.1007/s00603-012-
0220-7
Hantz, D., Arcamon, J., Laversanne, J.: Open–pit slopes back–analysis in complex rock
formation. In: Chengxiang, L., Ling, Y. (eds.) Proceedings of the International Symposium on
Engineering in Complex Rock. Science Press (1988). https://doi.org/10.1007/s10346-010-
0221-0
Hoek, E., Bray, W.J.: Rock slope engineering, 3rd edn. Taylor & Francis (1981). https://doi.org/
10.1007/s12594-016-0500-z
Janbu, N.: Application of composite slip circles for stability analysis. In: Proceeding of European
Conference on Stability of Earth Slope (1954). https://doi.org/10.1016/j.cageo.2011.10.024
Mohtrami, E., Jafari, A., Amini, A.: Stability analysis of slopes against combined circular–
toppling failure. Int. J. Rock Mech. Min. Sci. (2014). https://doi.org/10.1016/j.ijrmms.2014.
06.008
Morgenstern, N.R, Price, V.E.: The analysis of the stability of generalized slip surfaces. Int.
J. Rock Mech. Min. Sci. ResearchGate (1965). https://doi.org/10.1080/17486025.2011.
578674
Oztekin, B., Topal, T., Kolta, C.: Assessment of degradation and stability of a cut slope in
limestone, Ankara–Turkey. Eng. Geol. ResearchGate (2006). https://doi.org/10.1016/j.
enggeo.2005.11.012
Phase2 V9.0/Rocsceince.: A two−dimensional finite element analysis of soil and rock for
underground and excavated slopes, User’s manual, Rocscience INC. Springer, Heidelberg
(2014)
Priest, S.D., Brown, E.T.: Probabilistic stability analysis of variable rock slopes. Trans. Inst. Min.
Metall. Elsevier (1983). https://doi.org/10.1016/0148-9062(90)94333-O
Read, J., Stacey, P.: Guidelines for Open–Pit Slope Design, p. 512. CSIRO Publishing (2009)
RocLab 1.0/Rocscience: Program for determining rock mass strength parameters, Shear strength
of rock discontinuities, Rocscience Inc. Toronto, Ontario, Canada (2014)
Rosengren, K., Associates: Proposed Handlebar Hill open cut geotechnical review (Technical
Report No. 27010). Xstrata, Mt Isa, Australia (2007)
Rosengren, K., Associates.: Handlebar Hill open cut Pit. Internal technical report, in Xstrata
report in 2011, Xstrata, Mt Isa, Australia (2008)
Sjoberg, J.: Analysis of large scale rock slopes. Doctoral Thesis, Lulea University of Technology,
Lulea, Sweden (1999)
Skempton, A.W.: The (u = 0) analysis of stability and its theoretical basis. In: Proceeding of the
2nd International Conference of soil Mechanics and Foundation Engineering. Ice Virtual
Library (1948). https://doi.org/10.1680/geot.1948.1.1.1
Slide 6.0.: 2D limit equilibrium slope stability analysis. User’s manual, Rocscience. nrc research
press (2014). https://doi.org/10.1139/cgj-2014-0111
Spencer, E.: A method of analysis of the stability of embankments assuming parallel inter–slice
forces. J. Geotech. Sci. Educ. (1967). https://doi.org/10.12691/ajna-1-1-2
Terzaghi, K.: Theoretical Soil Mechanics, 1st edn. Wiley (1943). https://doi.org/10.1002/
9780470172766
Warren, N.S., Watters, J.R.: Limit equilibrium stability analysis utilizing geotechnical date at Mt.
Baker. ProQuest Dissertations Publishing, University of Nevada, Reno (2008). https://doi.org/
10.1139/cgj-2014-0111
Woldearegay, K., Riedmuller, G., Schubert, W., Mogessie, A.: Controlling parameters and
failure mechanisms of a large scale landslide in Ethiopia. FELSBAU 22(3) (2004)
Wyllie, D.C., Munn, F.J.: The use of movement monitoring to minimize production losses due to
pit slope failures. In: Proceedings, 1st International Symposium on Stability in Coal Mining
(1979)
Wyllie, D.C., Mah, C.W.: Rock Slope Engineering, Civil and Mining, 4th edn. Routledge (2004)
Reflection of Processes of Non-equilibrium
and Two-Phase Filtration in Fluid Saturated
Hierarchic Inclusion in a Block Layered
Medium by Data of Active Wave
Geophysical Monitoring
Olga Hachay1(&), Andrey Khachay2, and Oleg Khachay2

1
Institute of Geophysics Ural Branch of Russian Academy of Sciences,
Yekaterinburg, Russia
olgakhachay@yandex.ru
2
Ural Federal University, Yekaterinburg, Russia
Abstract. A new method for modeling acoustic monitoring of a layered-block

elastic medium with several inclusions of various physical-mechanical and
phase hierarchical structures was developed. An iterative process for solving the
direct problem is developed for the case of three hierarchical inclusions of l, m,
s-th ranks on the basis of the use of 2D integral and integral- differential
equations. The degree of hierarchy of inclusions is determined by the values of
their ranks, which can be different for each inclusion. Hierarchical inclusions are
located in different layers above each other: the upper is anomalously strained,
the second is fluid-saturated and the third is anomalously dense. The degree of
filling inclusions of each rank for all three hierarchical inclusions is different.
The results of the simulation can be used for monitoring of the fluid flow of oil
deposits.
Keywords: Hierarchical environment Acoustic field Iterative algorithm

Integral and integro-differential equations Direct problem
1 Introduction
The most important result of geomechanical and geodynamic studies of the past cen-
tury was the discovery of a close relationship between global geodynamic and local
geomechanical processes caused by mining operations, especially in tectonically active
zones. No less important result of the research was the conclusion about the funda-
mental role of the block-hierarchical structure of rocks and massifs for explaining the
existence of a wide range of nonlinear geomechanical effects and the emergence of
complex self-organizing geosystems. Hierarchical structure is typical for many sys-
tems, especially for the Earth’s lithosphere, where more than 30 hierarchical levels
from tectonic plates with a length of thousands of kilometers to individual mill grains
of millimeter size were identified by geophysical studies (Prangishvilli et al. 2001).
Thus, the earth’s crust is not a continuous medium, but a discrete block system and,

https://doi.org/10.1007/978-3-030-02032-3_3
32 O. Hachay et al.
like any synergetic discrete ensemble, has hierarchical and self-similar properties
(Kocharjan et al. 2003).
Possible reduction in economic indicators in the global oil and gas industry is due
to outdated and largely erroneous concepts on the structure of non-oil and gas fields in
the geological section. We can note several negative consequences in the economy of
the fuel and energy complex, caused by outdated and erroneous ideas. First, unjusti-
fiably large volumes of drilling wells in those sections of the fields where either there
are no deposits, or there are very poor deposits, the development of which is knowingly
unprofitable. Secondly, the various methods of influencing the reservoir with the aim of
increasing oil recovery in those areas where there is practically no oil only worsen the
economic indicators with erroneous ideas about the genesis and spatial structure of the
deposits. It is known that the volumes of injected liquid grow several times faster than
the volumes of incremental oil production. Thirdly, unreasonably high expenses for
drilling and pumping water not only reduce the overall profitability of hydrocarbon
production, but also significantly increase the level of environmental pollution, the
liquidation of which will require additional costs. Fourth, the erroneous concepts of the
origin and evolution of hydrocarbon deposits have led to such production technologies,
under which there is a very rapid depletion of the deposits and significantly reduced the
renewability of resources. All this unjustifiably quickly reduces economic indicators in
developed fields. The main elements of the new paradigm in the fuel and energy
complex are the search, development and technology of development of small-sized,
but extraordinarily rich in reserves of active oil and gas generation. The geosoliton
concept of the formation of hydrocarbons and the formation of large-scale, but rich in
reservoirs, is the main conceptual basis of the new paradigm (Bembel et al. 2004). To
date, new effective technologies can be attributed the results published in the paper
Levashov et al. (2017).
The processes of development of oil and gas fields are associated with the motion
of multiphase multicomponent media that are characterized by no equilibrium and non-
linear rheological properties. The real behavior of reservoir systems is determined by
the complexity of the rheology of moving fluids and the morphological structure of the
porous medium, as well as by the variety of processes of interaction between the liquid
and the porous medium (Khasanov et al. 2003). Accounting for these factors is nec-
essary for a meaningful description of filtration processes due to the nonlinearity, non-
equilibrium and heterogeneity inherent in real systems. In this case, new synergetic
effects are revealed (loss of stability with the appearance of oscillations, the formation
of ordered structures). This allows us to propose new methods for monitoring and
managing complex natural systems that are tuned to account for these phenomena.
Thus, the reservoir system from which to extract oil is a complex dynamic hierarchical
system.
When constructing a mathematical model of a real object, it is necessary to use, as a
priori information, active and passive monitoring data obtained during the current
operation of the facility. In papers of Hachay et al. (2011), (2013), modeling algorithms
were constructed for 3-D heterogeneities in electromagnetic case, for 2-D hetero-
geneities in the seismic case and for an arbitrary type of excitation source of an N-layer
Reflection of Processes of Non-equilibrium and Two-Phase Filtration 33
medium with hierarchical elastic inclusions located in the J-th layer. In the paper of
Hachay et al. (2014), a new 2D modeling algorithm for sound diffraction on elastic and
porous, fluid-saturated inclusions of hierarchical structure located in the J-th layer of an
N-layer elastic medium was developed. In Hachay et al. (2017) modeling algorithms
were constructed in the acoustic case for a 2-D heterogeneity and an arbitrary type of
excitation source of an N-layer medium with a separate hierarchical anomaly density,
stresses and plastic inclusions located in the J-th layer.
In this paper, using the method described in Hachay et al. (2011), (2013), (2014),
(2017), an algorithm for modeling the acoustic field (longitudinal acoustic wave) has
been developed in the form of an iterative process for solving a direct problem for the
case of three hierarchical inclusions of l, m, s-ranks using 2D integral and integro-
differential equations. The degree of hierarchy of inclusions is determined by the values
of their ranks, which can be different. Hierarchical inclusions are located in different
layers above each other: the top is anomalously stressed, the second is by fluid satu-
rated and the third is anomaly dense (Fig. 1).
Fig. 1. The scheme of composite anomalously stressed (upper), fluid-saturated (medium) and
anomalously denser (lower) heterogeneities of hierarchical type located in an N-layer elastic
medium.
34 O. Hachay et al.
1.1 Algorithm of Modeling of Sound Diffraction on 2D Composite

Heterogeneity of Hierarchical Type, Located in N-Layered Elastic
Environment
An algorithm for modeling of sound diffraction on a two-dimensional elastic hierar-
chical inclusions located in the J-th layer of an N-layer medium is described in Hachay
et al. (2013), (2017). GSp;j ðM; M 0 Þ- the source function of the seismic field, the
boundary value problem for which was formulated in Khachay (2006a, b), k12 ji ¼
x2 ðrji =kji Þ is the wave number for the longitudinal wave; in the above expression, the
index ji denotes the property of the medium inside the heterogeneity, ja denotes the
property of the medium outside the heterogeneity, k is the Lamé constant, r -density of
medium, x - circular frequency, ~ u ¼ grad u- vector of displacements, u0 - potential of
normal seismic field in layered medium in the absence of heterogeneity: u0ji ¼ u0ja . u0l -
potential of a normal seismic field in a layered medium in the absence of a hetero-
geneity of the previous rank l if l = 2 … L, if l = 1, which coincides with the corre-
sponding expression (Hachay et al. 2017). We assume that the density of the first
hierarchical inclusion for all ranks l and the enclosing layer are identical, and elastic
parameters of hierarchical inclusion for all ranks differ from the elastic parameters of
the enclosing environment. Let the rank values for all hierarchical inclusions will be:
l = m = s = 1, then the system of equations describing the propagation of the longi-
tudinal acoustic wave in the first inclusion will be rewritten as:
ðk1ðj1Þil
2
k1ðj1Þ
2
Þ ZZ
ul ðMÞGSp;ðj1Þ ðM; M 0 ÞdsM þ u0l1 ðM 0 Þ
2p
S1cl
¼ ul ðM Þ; M 2 S1Cl
0 0
ð1Þ
rðj1Þil ðk1ðj1Þil
2
k1ðj1Þ
2
Þ ZZ
ul ðMÞGSp;ðj1Þ ðM; M 0
ÞdsM þ u0l1 ðM 0 Þ
rðM 0 Þ2p
S1cl
¼ ul ðM 0 Þ; M 0 62 S1Cl ; M 0 2 Pj1
ZZ
rjil ðk1jil
2
k1j
2
Þ
ul ðMÞGSp;j ðM; M 0 ÞdsM þ u0l1 ðM 0 Þ ¼ ul ðM 0 Þ; M 0 2 Pj ð2Þ
rðM Þ2p0
S1cl
We calculate ul ðM 0 Þ; M 0 2 Pj in the layer where the second hierarchical elastic

inclusion is located, using expression (2), then the normal acoustic field potential for
the second inclusion is written in the form: u0m1 ðM 0 Þ ¼ ul ðM 0 Þ; M 0 2 Pj . The system
of equations for the second elastic hierarchical inclusion of rank m = 1 has the form
according to Hachay et al. (2011), (2013):
ZZ
ðk1jim
2
k1j
2
Þ rja 0
um ðMÞGSp;j ðM; M 0 ÞdsM þ u ðM 0 Þ
2p rjim m1
S2Cm
Z
ðrja rjim Þ @um
GSp;j dc ¼ um ðM 0 Þ; M 0 2 S2Cm ;
rjim 2p @n
C2m
2 ZZ
ð3Þ
rjim ðk1jim
2
k1j Þ
um ðMÞGSp;j ðM; M 0 ÞdsM þ u0m1 ðM 0 Þ
rðM Þ2p
0
S2Cm
Z
ðrja rjim Þ @u
GSp;j m dc ¼ um ðM 0 Þ; M 0 62 S2Cm ;
rðM Þ2p
0 @n
C2m
rðj þ 1Þim ðk1ðj ZZ

þ 1Þim k1ðj þ 1Þ Þ
2 2
um ðMÞGSp;ðj þ 1Þ ðM; M 0 ÞdsM þ u0m1 ðM 0 Þ
rðM 0 Þ2p
S2Cm
Z ð4Þ
ðrðj þ 1Þa rðj þ 1Þim Þ @u
GSp;ðj þ 1Þ m dc ¼ um ðM 0 Þ; M 0 62 S2Cm ; 2 Pj þ 1 ;
rðM 0 Þ2p @n
C2m
We calculate um ðM 0 Þ; M 0 62 S2Cm ; 2 Pj þ 1 in the layer where the third hierarchical

anomalous density inclusion is located using expression (4), then the normal potential
of the acoustic field is: u0s1 ðM 0 Þ ¼ um ðM 0 Þ; M 0 2 Pj þ 1 . We assume that the elastic
parameters of the third hierarchical inclusion for all ranks s and the enclosing layer are
the same, and the density of the hierarchical inclusion for all ranks differs from the
density of the host environment, then the system of equations for the third hierarchical
inclusion of rank s = 1 has the form according to Hachay et al. (2017):
ðk1ðj ZZ
þ 1Þis k1ðj þ 1Þ Þ
2 2
rðj þ 1Þa 0
us ðMÞGSp;ðj þ 1Þ ðM; M 0 ÞdsM þ u ðM 0 Þ
2p rðj þ 1Þis s1
S3Cs
Z
ðrðj þ 1Þa rðj þ 1Þis Þ @us
GSp;ðj þ 1Þ dc ¼ us ðM 0 Þ; M 0 2 S3Cs ;
rðj þ 1Þis 2p @n
C3s
ZZ ð5Þ
rðj þ 1Þis ðk1ðj
2
þ 1Þis k1ðj þ 1Þ Þ
2
us ðMÞGSp;ðj þ 1Þ ðM; M 0 ÞdsM þ u0s1 ðM 0 Þ
rðM 0 Þ2p
S3Cs
Z
ðrðj þ 1Þa rðj þ 1Þis Þ @us
GSp;ðj þ 1Þ dc ¼ us ðM 0 Þ; M 0 62 S3Cs 2 Pj þ 1 ;
rðM 0 Þ2p @n
Cm
GSp;ðj þ 1Þ ðM; M 0 Þ- function of the source of the seismic field, it coincides with the
þ 1Þis ¼ x ðrðj þ 1Þis =ðkðj þ 1Þis Þ; kðj þ 1Þis ¼ kðj þ 1Þa ; -
2 2
function (Khachay 2006a, b), k1ðj
wave number for the longitudinal wave and elastic parameters, in the above expression,
36 O. Hachay et al.
the index ji denotes the belonging of the medium properties inside the heterogeneity,
ja - outside the heterogeneity, s = 1, …, S - number of the hierarchical levels,
u0s1 ðM 0 Þ ¼ um ðM 0 Þ; M 0 2 Pj þ 1 is the potential of the normal acoustic field in the
layer j + 1 in the absence of the third heterogeneity of the previous rank. We calculate
us ðM 0 Þ; M 0 62 S3Cs 2 Pj1 in the layer j − 1 using expression (6):
rðj1Þis ðk1ðj1Þis
2
k1ðj1Þ
2
Þ ZZ
us ðMÞGSp;ðj1Þ ðM; M 0 ÞdsM þ u0s1 ðM 0 Þ
rðM 0 Þ2p
S3Cs
Z ð6Þ
ðrðj1Þa rðj1Þis Þ @u
GSp;ðj1Þ s dc ¼ us ðM 0 Þ; M 0 62 S3Cs 2 Pj1 ;
rðM 0 Þ2p @n
Cm
The values of L, M, and S are the maximum values of the ranks of the hierarchy for
the three inclusions. In this paper,
L ¼ 3; M ¼ 3; S ¼ 4: ð60 Þ
Let: l = l + 1; m = m + 1; s = s + 1. If l < 3 or l = 3, then we turn to the algorithm

(1)–(6). If l > 3, and m = 2, then we calculate us ðM 0 Þ; M 0 62 S3Cs 2 Pj in the layer
j using expression (7):
ZZ
rjis ðk1jis
2
k1j
2
Þ
us ðMÞGSp;j ðM; M 0 ÞdsM þ u0s1 ðM 0 Þ
rðM 0 Þ2p
S3Cs
Z ð7Þ
ðrja rjis Þ @us
GSp;ðj1Þ dc ¼ us ðM 0 Þ; M 0 62 S3Cs 2 Pj ;
rðM 0 Þ2p @n
Cm
u0m1 ðM 0 Þ ¼ us1 ðM 0 Þ; M 0 2 Pj and we proceed to the algorithm (3)–(6), if

m = 3, u0m1 ðM 0 Þ ¼ us1 ðM 0 Þ; M 0 2 Pj then we turn to the algorithm (3)–(6). If
m > 3, and s < 4 or s = 4, u0s1 ðM 0 Þ ¼ us1 ðM 0 Þ; M 0 2 Pj þ 1 , then we go to (5)–(6).
If s > 4, then we pass to (7′).
ZZ
ðk1jim
2
k1j
2
Þ rja 0
um ðMÞGSp;j ðM; M 0 ÞdsM þ u ðM 0 Þ
2p rjim m1
S2Cm
Z
ðrja rjim Þ @um
GSp;j dc ¼ ðum ðM 0 Þ þ ap2 Þ; M 0 2 S2Cm ;
rjim 2p @n
C2m
2 ZZ
ð70 Þ
rjim ðk1jim
2
k1j Þ
um ðMÞGSp;j ðM; M 0 ÞdsM þ u0m1 ðM 0 Þ
rðM 0 Þ2p
S2Cm
Z
ðrja rjim Þ @um
GSp;j dc ¼ um ðM 0 Þ; M 0 62 S2Cm ;
rðM 0 Þ2p @n
C2m
where a ¼ 1 v KK0 , K ¼ k-complete compression module, v-porosity, K0 -the true

modulus of phase compressibility, pore hydrostatic pressure p2. The first equation of
system (3) is rewritten in the form in accordance with the fact that in the second
inclusion the heterogeneities of the second rank become fluid or oil-saturated. Then go
to (5)–(6).
ZZ
rjis ðk1jis
2
k1j
2
Þ
us ðMÞGSp;j ðM; M 0 ÞdsM þ /0s1 ðM 0 Þ
rðM 0 Þ2p
S3Cs
Z ð8Þ
ðrja rjis Þ @us
GSp;j dc ¼ us ðM 0 Þ; M 0 62 S3Cs 2 Pj ;
rðM 0 Þ2p @n
Cm
We calculate us ðM 0 Þ; M 0 62 S3Cs 2 Pj in all the layers j = 1, … n with the use of

the expression (8). The algorithm stops if the hierarchy ranks become larger than the
given numbers (6′). If at some hierarchical level the structure of the local heterogeneity
breaks down into several heterogeneities, then the double and contour integrals in the
expressions (1–7′) are taken over all heterogeneities of a given rank.
2 Conclusions
Iterative modeling algorithms are constructed in the seismic case in the acoustic
approximation for a composite hierarchical heterogeneity. When constructing a com-
plex seismic gravity model without taking into account the anomalous influence of a
stress-strain state within a hierarchical inclusion, called the cushion of the deposit, an
analysis of the anomalous acoustic effect using data on the propagation of a longitu-
dinal wave shows that the influence of anomalous elastic parameters in the seismic
model cannot be neglected, since they affect on the values of the unknown anomalous
densities. If these values are used in constructing a density gravitational model without
taking into account the effect of elastic parameters, these density values will not reflect
the material composition of the analyzed medium. When constructing an anomalously
strained geomechanical model without taking into account the anomalous influence of
density heterogeneities inside the hierarchical inclusion, which is the substrate for a
two-phase deposit, the values of the unknown anomalous elastic parameters that cause
the anomalous stress state in the cushion using data on propagation of the transverse
wave will be determined incorrectly. These values of elastic parameters will not reflect
the real stress state of the analyzed medium over the fluid containing the deposit, which
in turn is represented by a hierarchical multi-neighborhood environment. For the first
time, the proposed iterative algorithm for modeling a hierarchically complex two-phase
medium can be used to manage the production of viscous oil in mine conditions and
light oil in sub horizontal wells (Khachay et al. 2016). The benefits of the new para-
digm, of the new methods of oil and gas mapping and monitoring using geophysical
wave fields on the base of developed mathematical algorithms can be seriously
enlarged.
38 O. Hachay et al.
References
Bembel, R.M., et al.: Geosoliton: The Functional System of the Earth, the Concept of
Exploration and Development of Hydrocarbon Deposits. Vector Book, Tyumen (2004)
Hachay, O.A., et al.: Integration of seismic and electromagnetic active methods for mapping and
monitoring the state of two-dimensional heterogeneities in an N-layer medium. Bull. South
Ural State Univ. Ser. Comput. Technol. Manage. Radio Electr. 2, 49–56 (2011)
Hachay, O.A., et al.: Modeling of electromagnetic and seismic fields in hierarchically
heterogeneous media. In: Bull. South Ural. State University. Ser. Comput. Math. Inform. 2
(2), 48–55 (2013)
Hachay, O.A., et al.: Reflection of no equilibrium two-phase filtration processes in oil saturated
hierarchical media in active wave geophysical monitoring data. Min. Inf. Anal. Bull. 4, 232–
238 (2014)
Hachay, O.A., et al.: Integration of acoustic, gravitational, and geomechanical fields in
hierarchical environments. Min. Inf. Anal. Bull. 4, 328–336 (2017)
Khachay, A.Y.: Algorithm for solving the direct dynamic seismic problem when excited by a
horizontal point force located in an arbitrary layer of an n-layer elastic isotropic medium. In:
Informatics and Mathematical Modeling, pp. 170–278. Ural State University, Ekaterinburg
(2006a)
Khachay, A.Y.: Algorithm for solving the direct dynamic seismic problem when excited by a
vertical point force located in an arbitrary layer of an n-layer elastic isotropic medium. In:
Informatics and Mathematical Modeling, pp. 279–310. Ural State University, Ekaterinburg
(2006b)
Khachay, Y., et al.: Effect of convective transport in porous media on the conductions of organic
matter maturation and generation of hydrocarbons in trap rocks complexes. Energy Procedia.
74, 79–83 (2016)
Khasanov, M.M., et al.: Nonlinear and no equilibrium effects in rheological complex
environments. Institute of Computer Research, Izhevsk, Moscow (2003)
Kocharjan, G.G., et al.: Dynamics of deformation of block massif. IKZ “Akademkniga”,
Moscow (2003)
Levashov, S., et al.: On the possibility of using mobile and direct-prospecting geophysical
technologies to assess the prospects of oil-gas content in deep horizons. In: Oil and Gas
exploration. Methods and Application, pp. 209–236. AGU, Willey, USA (2017)
Prangishvilli, I.V., et al.: System laws and regularities in electrodynamics, nature and society.
Nauka, Moscow (2001)
Evaluation of the Landslide in Erzurum
Ski-Jumping Complex - A Case Study
Murat Mollamahmutoğlu(&)
Civil Engineering Department, Engineering Faculty, Gazi University,

Ankara, Turkey
mmolla@gazi.edu.tr
Abstract. We present the assessment of a landslide that took place in the

Erzurum ski-jumping complex constructed for the 2011 Winter Universiade in
Turkey. The lower part of the ski jumps collapsed. Three of the jumps were
completely destroyed whilst the two larger jumps were severely damaged. To
begin with, the collapsed ski-jumping complex was inspected immediately and
periodically after the landslide. Thereafter, the related projects and site inves-
tigation report were obtained and examined in detail. Since the report for the site
was found to be insufficient, additional in-situ and laboratory tests were planned
and carried out to better understand the geotechnical characterization of the site
and compensate for the absence of data needed for the related geotechnical
analyses. Furthermore, the detailed slope stability analyses under different
conditions regarding various factors were conducted. All these studies revealed
that the landslide was not sudden but went through a creeping process. The
collapse of ski-runs in the landslide was due to the inadequate geotechnical
investigation and engineering judgment, design and application mistakes and the
lack of stability analyses and supervision.
1 Introduction
Landslides are a worldwide problem. They can lead to loss of life, loss of livestock and
damage the residential and industrial developments, villages or even entire towns,
destroying agricultural and forest lands. 2620 fatal landslides between 2004 and 2010
killed a total of 32,322 people excluding landslides triggered by earthquakes (Petley
2012).
The triggering mechanisms of natural slope failures frequently comprise a complex
interaction between hydrological and geotechnical processes, which in turn depends on
irregular topography, hydro-geotechnical properties, boundary conditions such as
permeability, and the initial state of the slope (Sorbino and Nicotera 2013). From a
geotechnical perspective, one of the main reasons for slope failure is the loss of matric
suction and hence a loss of effective stress as water infiltrates the soil.
Many researchers (Kim et al. 2004; Fredlund et al. 2012) have studied rainfall-
induced landslides through laboratory, field tests, and numerical analysis. They have
presented a mechanism for rainfall-induced landslides that commonly occur when the
wetting bands progress into the soil, resulting in a loss of suction and a reduction in
effective stress (Lu and Godt 2008).

https://doi.org/10.1007/978-3-030-02032-3_4
40 M. Mollamahmutoğlu
Geologic factors have also been found to cause mass movements on the slopes and
these include the presence of shallow soils over hard, impermeable rocks or glacial till,
soft, clay-rich rocks that produce thick plastic soil mantles, alignment of lithology
parallel to the ground slope and planar rock structures, unconsolidated or weakly
consolidated deposits (Sidle 1985).
Furthermore, landslides are commonly triggered by the modification of slope sta-
bility conditions, increasing stress, or reductions in strength (Rahardjo et al. 1988; Ng
et al. 2003, Ng et al. 2008). The landslide analysis on a large scale is usually performed
using deterministic methods such as the Limited Equilibrium Method (LEM) and Finite
Element Method (FEM). They are also useful in determining the factors involved and
the mechanism of slope failures, such as slope geometry, the mechanical properties,
and equilibrium condition.
In this paper, we present the results of the comprehensive investigation aimed at
determining the causes of the activation of a landslide in the Erzurum ski-jumping
complex, Turkey.
2 Study Area
2.1 Description
The Erzurum Ski Jump, or officially known as Türk Telekom Ski Jumping Towers, is a
ski jumping complex located on the Kiremitlik Hill at the base of Palandöken
Mountain, southwest of Erzurum in eastern Turkey. The complex, completed at the end
of 2010, consists of a large hill and a normal hill jumping tower, as well as three
additional smaller slopes of K-65, K-40, and K-20 for training purposes and for use by
young jumpers. The center with five ski jumping slopes covers an area of 30 ha. The
large hill has a K-point of 125 and a hill size of 140 with a 43.47 m high jumping
tower. It has a 99 m long inrun with an angle of 35°, a take-off angle of 11°, which is
7 m long and 3 m high. The slope of the landing is 34.57°. The total height is 136.8 m.
The normal hill’s K-point is 95 and hill size 109. It has a jumping tower of 20.98 m
height. The inrun is 88 m long at an angle of 35°. The take-off length is 6.5 m, the
height 2.5 m with an angle of 10.5°. The landing’s angle is 34.25°. The normal hill has
a total height of 107.1 m. The two Olympic-size towers have cylindrical structures at
the top. The venue has 10,000 spectator capacity (Fig. 1a).
During the time when no competitions are held, the facility attracts local population
and tourists for recreation activities and social events at the hotel restaurant and the
restaurant atop the 43 m-high tower, which offers a panoramic view of the entire city.
The whole complex cost about €20 million. The venue hosted fourth international
competitions, the last one being in 2012 and then the whole ski-runs was completely
damaged in a landslide in 2014 (Fig. 1b).
2.2 Regional Geology and Tectonic Setup

The study area is characterized by the presence of Pliocene age Gelinkaya Formation
deposited in lake and river environment. It is mainly composed of less-consolidated
Evaluation of the Landslide in Erzurum Ski-Jumping Complex 41
siltstone (Fig. 2a) sporadically interbedded with sandstone, gravels, and tuff (Figs. 2b,
3a, and b).
Fig. 1. Views of the completed and failed ski jump complex
Fig. 2. Close-up views of thinly layered siltstone and siltstone with coarse poorly sorted
materials
The strata in the Kiremitlik hill where the landslide occurred are highly fractured
and deformed. The faults on both sides of Kiremitlik hill are thought to be the source of
tectonic deformations but there is not sufficient information about them (Yılmaz and
Uysal 1988).
3 Methodology
The method of this research comprised an assessment of available data, surface and
subsurface investigations, in-situ and laboratory tests and the slope stability modeling
and analyses.
Fig. 3. Views of siltstone with less-consolidated sandstone layers and alternation of siltstone
and tuff
3.1 Assessment of Available Data

Geophysical, Geological and Geotechnical Studies
Surprisingly, there were a few boreholes within the slope that failed. Its soil
structure was inferred by means of some surrounding boreholes, which, however, had
insufficient depth of 10 to 12 m; none of them were extended to a depth of firm strata.
Similarly, the geophysical survey was also superficial reaching the depth of 20 m at
most. During borehole drilling, SPT was used to evaluate the strength of the layers
encountered, but the results were misleading due to the fact that most of the drilled
layers contained gravels and stones. Moreover, the position and orientation of the
layers in the landslide area were misinterpreted as horizontal whereas the later site
investigation revealed the presence of more complex structure. Thirteen new boreholes
whose depths ranged from 30 to 111 m, were driven to better understand the local soil
structure. The borehole locations are shown in Fig. 4. Since the access to failed area
was not possible they are scattered around wherever possible.
Related Projects
Cuts and fills were planned in the related projects to form standard slopes of ski-runs
(Fig. 5a). In addition, to protect skiers from the wind effect in the course of jumping
earthfill was designed next to the top of K125 and K65 ramps (Fig. 5b). But none of
these arrangements were based on the slope stability analyses.
A great mistake in the planning was the missing of the effective drainage system to
prevent the adverse effect of surface water infiltration into the ground (Fig. 5c).
Managerial Control and Consultancy Services
Supervision of construction activities on site was carried out jointly by the engineers
of the Ministry of Youth and Sport and the project firm. Nevertheless, it seemed that
they were not qualified in assessing the site investigation report and the related projects
and reading the conditions encountered on the site during construction activities.
Figure 6 showed that some local slope failures took place during field works.
Fig. 4. Boreholes layout
Fig. 5. Slope arrangements of K125, K65, and K40 in runs
Slope Creeping
The ski jump complex in question was completed at the end of 2010 and brought into
service in January 2011. After hosting the event of 2012 junior winter sport, heave
process took place at the lower part of K 125 ski-run concrete slab between the 80th
and 90th m towards the end of May 2012 (Fig. 7a). This was considered to be the slab
damage and repaired. In the first week of June 2014, on a rainy day, at the same place
of K125 ski-run concrete slab, another heave occurred extending to K65, K40, and K20
ski-runs’ concrete slabs (Fig. 7b). In the meanwhile, there were cracks above the
heaving line. The damaged parts were again repaired. Notwithstanding, a week later
after the repairment, heave and cracks repeated on the same lines. From that day on,
creeping of slopes increasingly continued. The process went on about five days and the
cracks on the upper part of K125 and K95 ski-runs were widened (Fig. 8a). At the
same time, on the other parts of the ski-runs undulating surfaces developed. Further-
more, the windbreaker earthfill near the top of K125 ski-run ruptured and the surface
deformations increased (Fig. 8b). The area was completely evacuated a day before the
failure of the ski-runs, that is on 15 July 2014.
Fig. 6. Local slope failures in the field during construction
Fig. 7. Creeping processes of K125, K65, K40 and K20 ski-runs
Fig. 8. Ruptures on the upper part of K125 ski-run and the windbreaker earthfill
It seemed that the slope deformation increments coincided with the periods of precip-
itation, watering, and snowing of the ski-runs for skiing. Thus, the penetration of surface
water into the ground accelerated the creep of the slopes and finally triggered the landslide.
3.2 Landslide Analysis

Landslide analyses were done considering both the original ground conditions and the
new conditions aroused from the cuts and fill to form the standard ski-runs of the
complex. Owing to the less-consolidated and highly fractured nature of Gelinkaya
formation, it was difficult to take undisturbed soil specimens during boring. However,
those managed to be taken, wherever possible, were subjected to relevant laboratory
tests to determine the geotechnical parameters essential to slope stability analyses. The
total unit weight, cohesion intercept, and the internal friction angle of the representative
samples varied from 19.1 to 20.2 kN/m3, 15 to 30 kPa, and 27 to 30° respectively. The
direct shear test was used to determine the shear strength parameters of the specimens
under drained condition since no groundwater was encountered up to 100 m of boring
depth. In addition, surface water which penetrated into the ground drained easily due to
the porous and fractured nature of local soil. Therefore, the geotechnical parameters
used in slope stability analysis were effective ones. Furthermore, the total unit weight,
cohesion intercept, and the internal friction angle of the earthfill material placed at the
top of the in-runs were found to be 20 kN/m3, 40 kPa, and 29° respectively. Slope
stability analyses regarding the most critical sections (K125 and K95 ski-runs) were
performed by Slide software and their results are presented under the following sub-
headings. The geotechnical parameters for each layer of slope profiles were given in
Table 1 as well as in the legends of Figs. 9, 10 and 11.
Table 1. The geotechnical parameters used in slope stability analyses
Pre-construction
The two most critical sections of the original ground condition of Erzurum ski-
jumping complex area was obtained from the related topography map. They were
where the K125 and K95 in-runs were constructed. In addition, the sections of slope
profiles were constructed in reference with data procured from the site reconnaissance,
in-situ, and laboratory test. The results of slope stability analyses by Slide software are
given below in Fig. 8. The factor of safeties against the slope failure of the virgin
ground under the static condition for both in-run areas were determined to be 1.424 and
1.484 respectively by modified Janbu method (Fig. 9).
Fig. 9. Slope stability analyses of the two most critical sections
Fig. 10. Slope stability analyses of the reshaped sections
Fig. 11. Back analyses of the reshaped sections
As shown, the virgin ground slope of the two area under consideration had no
instability problem at all.
Post-construction
To reshape the original slopes of the areas in question for the K125 and K95 ski-runs,
the original slopes were steepened by cut-and-fill and their weights were increased by
the surcharges due to earthfills. In fact, these processes brought the original slopes to a
limit equilibrium condition under the static condition as analyzed by modified Janbu
method (Fig. 10).
Back analysis has been proven to be one of a useful tool for the investigation of
slope failures. It determines the shear strength parameters needed to generate the
desired factor of safety. This analysis need the data such as geometry of slope debris,
pore water pressure, density of failure material and the strength of material along the
slip surface. There are several methods to do back analysis. However, in this case-
study, the shear strength parameters, namely cohesion intercept and friction angle
which resulted in a safety factor of 1 were searched by trial and error for K125 and K95
in-runs’ slopes to compare between the input data for the back analysis and the situ-
ation observed in the field including laboratory data. For the rearranged slope condi-
tions, the shear strength parameters determined in this way (Fig. 11 and Table 2) were
compared with those used in the previous analyses (Fig. 10 and Table 1); they were
found to be very close to each other.
Table 2. Comparison of shear strength parameters obtained from laboratory tests and back
analyses for the soil profiles under consideration
The above analyses revealed that the depth of failure surface was about 25 m and
the slope failure included some part of the earthfill at the top of the slopes. All these
were compatible with the actual situation encountered in the field.
4 Conclusions
1. Geological, geophysical and geotechnical data and their evaluation were quite
insufficient.
2. Slope stability analyses of the original ground and the rearranged conditions formed
by cuts and fills were not taken into account.
3. Project planning lacked in the effective drainage system, especially including the
five ski-runs’ area.
4. This case story could illustrate how important and cost-effective are comprehensive
geotechnical engineering site investigation before the construction of infrastructure
in risk areas.
5. To avoid geotechnical risk and economic losses localization of this type of sport
infrastructure could be carefully planned in advance and included detailed recog-
nition of geotechnical engineering condition.
6. Same of the areas could not be suitable for this type of infrastructure at all or cost of
remediation works could be too high.
References
Petley, D.: Global patterns of loss of life from landslides. Geology 40, 927–930 (2012)
Sorbino, G., Nicotera, M.V.: Unsaturated soil mechanics in rainfall-induced flow landslides. Eng.
Geol. 165, 105–132 (2013)
Kim, J., Jeong, S., Park, S., Sharma, J.: Influence of rainfall-induced wetting on the stability of
slopes in weathered soils. Eng. Geol. 75, 251–262 (2004)
Fredlund, D.G., Rahardjo, H., Fredlund, M.D.: Unsaturated Soil Mechanics in Engineering
Practice. Wiley, New York, USA (2012)
Lu, N., Godt, J.: Infinite slope stability under steady unsaturated seepage conditions. Water
Resource. Res. 44, 11 (2008)
Sidle, R.C., Pearce, A.J., O’Loughlin, C.L.: Hillslope stability and land use. Water Resources
Monograph. vol. 11. The American Geophysical Union. Washington, DC. 11, 140 (1985)
Rahardjo, H., Leong, E.C., Gasmo, J.M., Tang, S.K.: Assessment of rainfall effects on the
stability of residual soil slopes. In Proceedings of the 2nd International Conference on
Unsaturated Soils, vol. 2, pp. 280–285. Beijing, China (1988)
Ng, C.W.W., Zhan, L.T., Bao, C.G., Fredlund, D.G., Gong, B.W.: Performance of an unsaturated
expansive soil slope subjected to artificial rainfall infiltration. Géotechnique 53, 143–157
(2003)
Ng, C.W.W., Springman, S.M., Alonso, E.E.: Monitoring the performance of unsaturated soil
slopes. Geotech. Geol. Eng. 26, 799–816 (2008)
Yılmaz, A., Uysal, Ş.: 1:100 000 scale- geological mapping of Turkey – Map section F32. MTA
publication. Ankara. 20 p. (1988)
Geological Investigation and Risk Assessment
for Disaster Management of Merapi Volcano
and Surrounding Area, Yogyakarta Special
Territory, Indonesia
Ayu Narwastu Ciptahening1(&), Nandra Eko Nugroho1,

and Noppadol Phienwej2
1
Universitas Pembangunan Nasional “Veteran” Yogyakarta,
Yogyakarta, Indonesia
ciptaheningayu@gmail.com
2
Asian Institute of Technology, Khlong Luang, Thailand
Abstract. Mount Merapi is classified as one of the most attractive volcanos in

the world. It is situated in Central Java and Yogyakarta Special Territory,
Indonesia. It regularly erupts in every 4 years. Deposits produced by the activity
of Mount Merapi are spread out in Sleman Regency Yogyakarta Special Terri-
tory, where the study area is located. The area includes four rock units, specifi-
cally Ancient Merapi, Middle Merapi, Recent Merapi, and Modern Merapi.
Mount Merapi eruptions had been responsible of the collapse of Hindu
Mataram Kingdom, Central Java, in the 10th to 11th century periods. This was
demonstrated by the discovery of many ancient temples in Yogyakarta and
surrounding areas, such as Temple Sambisari, Temple Kadisoka, and Temple
Kedulan which were buried by volcanic deposits of Mount Merapi. The eruption
of Mount Merapi, in the year 2010, recognized as the biggest eruption in recent
decades, claimed about 300 lives and very large losses. The type of eruption
changed from Merapi type in the activity period of 2006 and before to become
the combination of Vulcanian type and Pelean type. Due to the very intensive
activities of the volcano, the geohazard threats faced by people living in the
surrounding area of Yogyakarta are high. And in the face of the volcanic dis-
asters, the people of Mount Merapi unite in the implementation of disaster
preparedness.
In the study area, the geohazards are represented by lava flow, pyroclastic
flow, ash fall, and lahar. There are four areas can reached by pyroclastic flow,
including the Turi, Pakem, Cangkringan, and Ngemplak districts. All of the
districts in the study area are threatened by ash fall, such as the Turi, Pakem,
Cangkringan, Tempel, Sleman, Ngaglik, Ngemplak, Minggir, Sayegan, Mlati,
Gamping, Godean, Depok, and Kalasan districts. While the areas threatened by
lahar are Turi, Pakem, Cangkringan, Tempel, Ngaglik, Ngemplak, and Kalasan
districts. By applying a scoring method, the study area can be classified into
zone with high, moderate, and low geohazard vulnerability, threat, and risk.
Keywords: Volcano Mount Merapi Geohazards Vulnerability

Threat Risk

https://doi.org/10.1007/978-3-030-02032-3_5
50 A. N. Ciptahening et al.
1 Introduction
1.1 Background
Indonesia is rich in natural resources and culture, but also suffers from many natural
disasters. The presence of geohazards is related to the tectonic setting of Indonesia
located at the juncture of three big lithosphere plates, i.e. Indo-Australia plate in the
south, Eurasia plate in the north, and Pacific plate in the east. Thus Indonesia belongs to
a seismically and volcanically active region known as the “ring of fire”. Indeed,
Indonesia has many active volcanoes.
One of the most active volcanoes in Indonesia is Mount Merapi located at the
northern part of Yogyakarta Special Territory. The eruption of Mount Merapi in the
year 2010 had made headlines around the world either through mass media or print
media. This was the biggest eruption in recent decades and it claimed lives and very
large losses. In the historical records, the most dangerous and destructive threats of
Mount Merapi activities are pyroclastic flows and debris/mud flows (lahars). There are
many changes that occur on geological aspects, benefits, and losses caused by the
eruption of Mount Merapi 2010. Due to the existence of Mount Merapi that is very
active, it can be underlined that the geohazards faced by people living in the sur-
rounding area of Yogyakarta is about the impact of volcanic eruptions.
Based on the above backgrounds, it is necessary to conduct Geological Investi-
gation, Risk Assessment and Cultural Behaviour Analysis for Disaster Management of
Merapi Volcano and Surrounding Area.
1.2 Objectives
The objectives of geological investigation in the Mount Merapi and surrounding area
are to study on geology and risk assessment for identifying the threat types and
mapping the geohazard distribution in the study area and performing a risk assessment
to determine the anticipation and prevention efforts in order to develop disaster
management.
1.3 Location and Accessibility of the Study Area

The study area is located in the Sleman regency, Yogyakarta Special Territory,
Indonesia. It is bordered by Magelang regency in the West, Boyolali regency in the
North, Klaten regency in the East, and Bantul regency in the south, with the total area
of 32 km x 32 km. Geographically, the study area lies between latitudes 7o 33’
05.8’’–7o 46’ 41.0’’ South latitude and longitude 110o 16’ 16.2’’–110o 29’ 18.4’’ East
longitude, and astronomically, the study area is located in 419595 mE–443595 mE and
9140205mN–9165205mN WGS 1984 UTM (Universal Transverse Mercator) (Fig. 1).
Geological Investigation and Risk Assessment 51
Fig. 1. Location of the study area
2 Method of Study
In this research carried out analytical and descriptive approaches, by processing data
both primary and secondary. The research methods applied in this study includes:
a. Literature review is done by studying the existing references of previous researchers
regarding the study area.
b. Collecting of secondary data includes the results of previous researches on the
geomorphology, geology (stratigraphy and geological structures), and geohazards.
c. Interpretation and analysis of topographic map and remote sensing. The map used
for the interpretation is a topographic map of 1: 75,000 scale, aerial photographs of
1: 30,000 scale, as well as satellite imagery.
d. Geological survey, mapping, and rock sampling. Field survey is done to check the
validity of some secondary data, including geological conditions, the spreading of
lithology and geological structure in the study area. On the other hand, geological
mapping itself is implemented from field data collection and rock sampling. The
field works are included conducting path, observation, measurements, sampling,
data documentation, recording and plotting of data into the field map.
e. Laboratory testing and analyses represented Petrologic testing and thin section
analyses are done to determine the type and the name of rocks.
f. Risk analysis and assessment is done as a basic of the preparation for risk reduction
and anticipation to the geohazard threat that may occur.
3 Result and Discussion
3.1 Geology
3.1.1 Geomorphologic Units in the Study Area
Morphology of a volcano is the result of interaction between endogenous and
exogenous processes. The morphology of a volcano is not only developed by the
material released and the type of eruption, but also by the degree of volcanic activity,
erosion, as well as the geologic structure. The study area is included in volcanic
landform, and divided to four geomorphologic units, they are:
Parasitic Cone (V1)

This unit is developed by volcanic process. It occupies 5% of the entire study area. It is
a volcanic hill with steep to very steep slope which is 21%–140% and has a range of
elevation 1000–1400 m above sea level (asl). Lithology occupying Parasitic Cone unit
are lava, autoclastic brecciated lava, and pyroclastic breccia. This unit is used for
cultivation of tree planting area and for foresting.
Upper Volcanic Slope (V2)

This unit is formed and influenced by volcanic process, occupying 10% of the entire
study area, with steep to very steep slope which is 21%–140% and has a range of
elevation 1350–2600 m asl. Upper Volcanic Slope unit is occupied by lava flow
deposits, pyroclastic breccia, laharic breccia, pyroclastic and laharic deposits. This unit
is usually used for cultivation area and farming land.
Middle Volcanic Slope (V3)

This unit is also formed and influenced by volcanic process, occupying 15% of the
entire study area, with steep slope which is 21%–55% and has a range of elevation
600–1350 m asl. Middle Volcanic Slope unit is occupied by pyroclastic breccia, laharic
breccia, fine grained pyroclastic deposits as sandstone, laharic deposit of unconsoli-
dated boulders, gravels, and sand. The unit is mostly used for agricultural area. There
are some villages located in this area.
Lower Volcanic Slope (V4)

This unit is formed mostly by volcanic deposition. It occupies 70% of the entire study
area, with sloping slope which is 8%–13% and has a range of elevation 150–600 m asl.
Lithology occupying the unit are generally unconsolidated pyroclastic and laharic
deposits. Lower Volcanic Slope unit is in general used for agricultural and residential
areas. There are many villages and town located in this unit (Fig. 2).
3.1.2 Rock Units/Litology of the Study Area

According to Camus et al. (2000), there are four periods of Merapi volcanic activities,
comprising Ancient Merapi period, Middle Merapi period, Recent Merapi period, and
Modern Merapi period.
In general, lithology of the study area is dominated by volcanic deposits of Mount
Merapi that can be grouped into the following rock units; Unit of Turgo Plawangan
Lava Deposit, Unit of Ancient Merapi Pyroclastic Deposit, Unit of Ancient Merapi
Laharic Deposit, Unit of Middle Merapi Lava Deposit, Unit of Middle Merapi
Fig. 2. Geomorphology map of the study area
Pyroclastic Deposit, Unit of Recent Merapi Lava Deposit, Unit of Recent Merapi
Pyroclastic Deposit, Unit of Recent Merapi Laharic Deposit, Unit of Modern Merapi
2006 Pyroclastic Deposit, Unit of Modern Merapi 2006 Lahar Deposit, Unit of Modern
Merapi 2010 Pyroclastic Deposit, Unit of Modern Merapi 2010 Lahar Deposit and Unit
of Alluvial Deposit.
The study area is geologically controlled by volcanic activity. The constituent rocks
are composed of Ancient Merapi deposit, Middle Merapi deposit, Recent Merapi
deposit, and Modern Merapi deposit.
Ancient Merapi Deposit

Ancient Merapi period is 14,000 to 40,000 years old. It produced lava, auto breccias,
pyroclastic and laharic breccias. The composition of the deposits is basaltic andesite.
Middle Merapi Deposit

Middle Merapi period produced thick lava deposit and pyroclastic breccias. The lava
can be obtained at Batulawang and Gajahmungkur hills, with andesitic composition.
Age of Middle Merapi period is 2,200 to 14,000 years. In the study area, the deposit is
represented by pyroclastic breccias, found in the northern part of Pakem and Cangk-
ringan districts.
Recent Merapi Deposit

Recent Merapi period produced deposits consists of thin lava (less than 3 m), pyro-
clastic flow and laha deposits of 1,000 to 2,000 years age. Recent Merapi shows a
particular character of activity called Merapi Type. It specifies with lava dome for-
mation, followed by lava avalanches forming pyroclastic flows and glowing clouds.
The deposit consists of lava, pyroclastic, tuffaceous sand and lahar that generally
unlithified yet, with andesitic composition.
Modern Merapi Deposit

The product of Mount Merapi that is less than 1000 years old is named Modern Merapi
Deposit. In this study, Modern Merapi Deposit is especially represented by the prod-
ucts of Merapi eruption 2006 and Merapi eruption 2010. Mineralogic composition of
Modern Merapi deposits is more acidic than those of previous periods. According to
Hammer et al. (2000), it categorized as high K-basaltic andesite (Fig. 3).
Fig. 3. Geologic map of the study area
Merapi 2006
Physically, the deposits of Merapi 2006 were resulted both from primary and secondary
activities, composed of pyroclastic flow, surge, and lava, with grain size of bolder,
sand, ash, and dust
Merapi 2010
Mount Merapi 2010 produced pyroclastic flow which is deposited more than 15 km
distant from the center of the eruption to the southeast direction through River Gendol
valley. There were 100 million m3 volcanic materials deposited on the upper slopes of
the volcano.
Alluvial Deposits
Alluvial deposits are especially composed of volcanic materials that weathered, eroded,
transported through river and finally deposited. This unit is found in the southern part
of the study area, in form of floodplains and bars.
3.2 Geohazards in the Study Area

In the study area, geohazards are related to volcanic activities, therefore represent
volcanic hazards including lava flow, pyroclastic flow, ash fall, and lahar.
3.2.1 Lava Flow

The lava of Mount Merapi is relatively viscous (Kusumayudha 2013), the temperature
varies from 800 °C to 1000 °C. Due to its viscosity, lava flow of Merapi doesn’t move
fast. Its flow distance reaches maximally 1,5 to 4,5 km from the eruption center, thus
people very rare killed by it, even though it is extremely hot (800 oC–1000 oC). The
petrologic composition of Merapi lava is andesitic to basaltic andesite. By interpreta-
tion on topographic map which is delineation from contours patterns, in the study area,
there are three districts that are threatened by lava flow. Those are the Turi, Pakem, and
Cangkringan distict. According to the records of Goelogical Survey of Indonesia, the
farthest of lava flow reached 4.5 km in the Pakem district.
3.2.2 Pyroclastic Flow

The most hazardous of Mount Merapi volcanic hazards is pyroclastic flow or glowing
cloud. Pyroclastic cloud with temperatures between 600 oC–800 oC is able to burn all
living creatures and various objects. The velocity is very high, up to 120 km/h resulted
in any difficulty to escape from its attack. The composition of pyroclastic flow is tuff
(ash), lapilli, volcanic sands, bombs, and blocks; is able to reach more than 15 km
distant from the eruption center. By interpretation using satellite image and field work,
there are four districts that able to be reached by pyroclastic flow, including the Turi,
Pakem, Cangkringan, and Ngemplak district (Fig. 4).
Fig. 4. Volcanic hazard map of the study area

3.2.3 Ash Fall

Ash fall belongs to an important threat of volcanic eruptions on the surrounding
environment, because it can cause respiratory disease to humans, damaging crops,
disrupting traffic in urban traffic including air flights, damaging infrastructure and
vehicles. The ash of Merapi contains volcanic tuff and fine sands, with mineral com-
position of dominantly plagioclase, amphibole, and silica (SiO2). During eruption, ash
fall can cover all districts in the study area, in another word, all of the study area is
threatened by this hazards, such as the Turi, Pakem, Cangkringan, Tempel, Sleman,
Ngaglik, Ngemplak, Minggir, Sayegan, Mlati, Gamping, Godean, Depok, and Kalasan
district.
3.2.4 Lahar
In general, it occurs in rainy season. Lahar is composed of water and materials with
various grain size from sand to very big boulder. The density of lahar is very high,
flowing lead by gravitational force; the speed of flowing reaches 60 km/h. The districts
affected by lahar are the Turi, Pakem, Cangkringan, Tempel, Ngaglik, Ngemplak,
Mingggir, Depok, and Kalasan district.
3.3 Volcanic Hazard Threatening Level

As it has been mentioned above, there are four kinds of geohazards exist in the study
area, including lava flow, glowing clouds or pyroclastic flow, ash fall, and mud flow or
lahar. Qualitatively the volcanic hazard threatening level can be written as follows
(Table 1):
Table 1. Volcanic hazard threatening level
After have been analyzed, the study area can divided into four threatening level.
The first is very high risk that threatened by lava flow, pyroclastic flow, ash fall and
lahar, including the Turi, Tempel, and Cangkringan distict; The second is high risk with
threats of pyroclastic flow, ash fall, and lahar, covering the Ngemplak district; The third
is moderate risk with threats of ash fall and lahar, including the Tempel, Depok, and
Kalasan dictrict; And the fourth is low risk that only in danger by ash fall, involving the
Sleman, Ngaglik, Sayegan, Minggir, Mlati, Godean, and Gamping district (Fig. 5).
Fig. 5. Summary of volcanic hazards threatening level
3.4 Vulnerability of the Study Area

Vulnerability is situations or conditions which are cause inability of the people to
prepare themselves to face natural disaster. Vulnerability covers very diverse. The
general situations that include factors of physical, socio-economy, political, and cul-
tural, which are could potentially give effect to the people to more easily affected by
disasters. The vulnerability in the study area can be classified in three vulnerability
classification, those are high vulnerability, moderate vulnerability, and low vulnera-
bility. To classify vulnerability in the study area, there are four main factors looked up;
Human, Infra-structure, Agriculture, and Socio ecomony (cost losses in Rupiah).
3.5 Risk Assessment

Disaster risk assessment is an approach to show potential negative impacts that may
arise as a result of a potential disaster struck. Potential negative impacts are calculated
based on the level of vulnerability and capacity of the region. The potential negative
impact is seen from the potential number of people who are exposed to, loss property
and environmental damage (Fig. 6).
Fig. 6. Result and Vulnerability Map of the study area

3.6 Threat Index

Threat determination of volcanic disaster viewed using Disaster-prone Areas map of
Center for Volcanology and Geological Hazard. By a scoring analysis, the result shown
in figure below (Fig. 7).
Fig. 7. Result and Threat Map of the study area
3.7 Disaster Risk Level

Disaster Risk level is determined by combining the Level Capacity Loss Rate. Disaster
Risk Level Determination conducted for each potential hazard in an area. By a scoring
analysis, the result shown in the figure below (Fig. 8).
Fig. 8. Results and disaster risk map of the study area

4 Conclusions
Based on the research and analysis that has been done in research, as well as the
discussion in the previous chapters, it can be concluded as follows:
1. Geohazards are natural occurrence that involves geological processes. In the study
area, geohazards that occur are related to volcanic activities, therefore it can be
named volcanic hazards that consist of lava flow, pyroclastic flow, ash fall, and
lahar. There are three districts that are threatened by lava flow, those are the Turi,
Pakem, and Cangkringan distict. There are four districts that able to be reached by
pyroclastic flow, including the Turi, Pakem, Cangkringan, and Ngemplak district.
All of the study area is threatened by ash fall, such as the Turi, Pakem, Cangk-
ringan, Tempel, Sleman, Ngaglik, Ngemplak, Minggir, Sayegan, Mlati, Gamping,
Godean, Depok, and Kalasan district. While areas which threatened by lahar are the
Turi, Pakem, Cangkringan, Tempel, Ngaglik, Ngemplak, and Kalasan district.
2. By applying scoring method on geohazard vulnerability, the study area can be
classified into three vulnerability classification, those are high vulnerability, mod-
erate vulnerability, and low vulnerability.
3. Based on volcanic hazard threatening analysis have been done, the study area can
divided into three volcanic hazard threatening classification. The first is high that
including Hargo Binangun village in the Pakem district and Glagah Harjo, Umbul
Harjo, Kepuh Harjo village in Ngemplak district. The second is moderate that
covering the Giri Kerto, Wono Kerto village in the Turi district, Purwobiangin
village in Pakem district and Wukirsari village in Cangkringan district. The third is
low that including the Mardirejo village in Tempel district, Binangun Kerto and
Dono Kerto village in Turi district, Candi Binangun, Pakem Binangun, Harjo
Binangun in Pakem district, and Argomulyo village in Cangkringan district.
4. Accordung to risk analysis, the study area included in high, moderate, and low risk
area. The high risk area covered Hargo Binangun village of the Turi district, and
Glagah Harjo, Umbul Harjo, Kepuh Harjo village of Ngemplak district. The
moderate risk area located in the Mardiko Rejo, Lumbung Rejo village of Tempel
district, Giri Kerto, Wono Kerto, Bangun Kerto, Dono Kerto village of Turi district,
Purwo Binangun, Candi Binangun, Pakem Binangun, Harjo Binangun village of
Pakem district, Wukir Sari, Argo Mulyo village of Cangkringan district, Suko
Harjo, Donoharjo village of Ngaglik district, and Umbulmartani, Bimo Martani,
Sindumartani village of Ngemplak district.
References
Kusumayudha, S.B.: Gunungapi Aktif di Indonesia (Active volcanoes of Indonesia), 2nd edn. PT
Citra Aji Parama, Yogyakarta (2013)
Toyo Santoso Dipo, H.: Bencana Tanah Longsor Kabupaten Kulon Progo dan Upaya Mitigasi
Bencana, Pros. Simposium Nasional Pencegahan Bencana Sedimen, pp. 40–49 (2002)
Van Bemmelen, R.W.: The Geology of Indonesia, vol. IA, 732 pp. Martinus Nijhoff, The Haque
(1949)
Acquisition and Analysis of Surface Wave Data
in the Indo Gangetic Basin
P. Anbazhagan1,2(&), Ketan Bajaj1, Sayed S. R. Moustafa2,3,

and Nassir S. N. Al-Arifi2
1
Department of Civil Engineering, Indian Institute of Science, Bangalore, India
anbazhagan@iisc.ac.in
2
Geology and Geophysics Department, Faculty of Science, King Saud
University, Riyadh 11451, Saudi Arabia
3
Seismology Department, National Research Institute of Astronomy and
Geophysics (NRIAG), Cairo 11421, Egypt
Abstract. Seismic activity in the tectonically active Himalayan region can

result in devastation in the contagious deep Indo-Gangetic Basin (IGB). It is
filled by alluvial deposits and, therefore, prone to site amplification. IGB had
experienced catastrophic earthquake damages due to the presence of thick soil
depth of 0.05 km to about 4 km. However, very few studies have presented
seismic site characterization of the IGB soil up to the shallow depth, and very
limited attempts have been made to measure the dynamic properties of the deep
soil column. Hence, in this study, shear velocity profile (Vs ) up to 400 m depth
is measured using combined active and passive multichannel analysis of surface
wave (MASW) survey in 275 selected locations in IGB. Acquisition and pro-
cessing of the seismic surface wave data to obtain the Vs profiles at shallow and
deep depths are discussed in this paper. Parametric study has been done to
identify the variation in dispersion and Vs profiles by recording time, different
sampling interval, different polarity and gain. It has been seen that changing any
parameter during data recording may result in variation in Vs . Further, the
surveyed sites are classified and characterized based on time-averaged Vs in the
upper 30 m depth as per National Earthquake Hazards Reduction Program
seismic site classification. Vs30 values vary from 157 to 1152 m/s in the entire
stretch of IGB. This is the first time such extensive study was carried out in IGB
for Vs determination at deep and shallow depths.
Keywords: Indo-Gangetic Basin MASW Data acquisition

Seismic site characterization Passive data recording
1 Introduction
Local site conditions can have great influence on ground surface motion and structural
damage by earthquakes. Two classic examples that emphasize the influence of site
amplification due to local site effect are the 1985 Mexico and the 1989 Loma Prieta
earthquakes. Similarly, many earthquakes in India (1934 Bihar-Nepal; 2001, Bhuj;
2015 Nepal earthquake) have also illustrated the importance of local site effects. The
Indian subcontinent is one of the most seismically active regions in the world.

https://doi.org/10.1007/978-3-030-02032-3_6
Acquisition and Analysis of Surface Wave Data 61
The large strains are building along and within the plate boundary due to the ongoing
collision between the Indian and Eurasian Plates. Moreover, crustal shortening
increases the earthquake hazard, particularly in the northern part of the Indian Sub-
continent. Various authors (Bilham and Ambraseys 2005; Bilham et al. 2001) have
studied the seismotectonics of the Himalayan region and predicted the high seismicity
along the entire stretch of the Main Boundary Thrust (MBT), the Main Central Thrust
(MCT) and Indus-Tsangpo Suture (ITS). In the last two centuries, the Himalayan
region has experienced many events of magnitude more than 7. The highly fertile and
deep basin of Indo-Gangetic Basin (IGB), bound on the north side by the Himalayas, is
one of the most populous areas. It is about 1000 km long to the south and filled-up by
loose soil deposits. High seismicity region surrounds the IGB and this makes the
damage scenario more destructive. Significant causalities to human life or infrastructure
loss can result from any large earthquake in the future. Hence, there is a need to study
the spatial variation of shear wave velocity in the IGB for determining the local site
effect due to the presence of deep deposits contagious with high seismicity region. The
proper characterization of deep soil deposits in the IGB also sets the priority for the
determination of seismic site classification and the amplification factor for different
periods due to local site effect.
Various authors (Boominathan et al. 2008; Anbazhagan and Sitharam 2008;
Anbazhagan et al. 2010; Kumar et al. 2012; Desai and Choudhury 2014, 2015; Kumar
et al. 2016) in India have attempted to determine the shear wave velocity at shallow
depths in different study areas. Anbazhagan and Sitharam (2008), determined the shear
wave velocity profile for shallow depth in Bengaluru. Desai and Choudhury (2015) and
Kumar et al. (2012) respectively derived the Vs profile for Mumbai and Lucknow
region in India. Jishnu et al. (2013) derived the Vs profiles at four different locations for
Kanpur for site response study. Pandey et al. (2016) derived the Vs profile of seismic
stations in Delhi region by carrying the multichannel analysis of seismic waves. Most
of the previous studies were limited to soil column of 30 m depth. Additionally, in
many of the studies, the measured SPT-N values were converted to Vs profiles and used
for site response studies. Till today there are no studies available for determining the
local site effect for the deep deposits of IGB from the measured Vs profiles exceeding
100 m depth.
There are several approaches for estimating Vs through surface wave analysis and
with the developments in geophysical-indirect methods, the degree of uncertainty is
reducing eventually. These geophysical methodologies integrate border soil volume
and, therefore, are more representative of the seismic behavior of a site than local in-
hole measurements (Humire et al. 2015). In this study, combined active and passive
multichannel analysis of surface wave (MASW) survey has been used for determining
the shear wave velocity profile at 275 locations in the IGB. For evaluating the dis-
persion characteristic both active and passive sources, linear and circular arrays have
been used. Geophones of 2 Hz frequency have been used for performing the MASW
survey. Ambient noise has been used as a source for both passive remote and roadside
survey. Recordings have been done at different sampling intervals and record lengths.
Measurements are performed at different locations in the entire stretch of IGB,
stretching across the Indian states from Bihar to Punjab. The sites are classified and
62 P. Anbazhagan et al.
characterized based on time-averaged Vs in the upper 30 m depth as per NEHRP

(BSSC 2003) seismic site classification.
2 Study Area
The IGB is shaped as a result of post-collision between Indian and Asian tectonic plates
(Singh 1996). It is a peripheral foreland basin system, formed on a flexed Indian plate
lithosphere (Lyon-Caen and Molnar 1985; Singh 1987). The Ganga Plain spreads from
Aravalli-Delhi Ridge to Rajmahal hills along the west and the east respectively;
Himalayan foothills in the north to the Bundelkhand Vindhyan Plateau in the south,
occupying an approximate area of 250,000 km2, lies between 77°E to 88°E (longitude)
and 24°N and 30°N (latitude). The length of IGB is about 1000 km; with the variable
width, ranging from 200 to 450 km, being wider and narrower, along the western and
the eastern part (See Fig. 1). The IGB is irregular along with outcrops of rocks
pounding out of the alluvial in the southern edge, whereas, the Siwalik rocks are
exposed in the northern margin beside the Himalayan Frontal Fault (HFF) (See Fig. 1).
The IGB shows all the foremost components of a foreland basin system, namely an
orogen (the Himalaya), deformed foreland basin deposits adjacent to the orogen
(Siwalik Hills), a depositional basin (Ganga Plain) and peripheral cratonic bulge
(Bundelkhand Plateau) (Singh 1996). IGB established in the Early Miocene, and from
Middle Miocene to Middle Pleistocene, the northern part of the IGB was uplifted, and
thrust basin-wards; the Ganga plains moved southwards because of thrust loading in
the orogen (Singh 1996).
Based on the geophysical study and deep drilling carried out by ONGC, various
researchers (Rao 1973; Lyon-Caen and Molnar 1985) depicted the basement structure
and sub-surface geology. As per, Rao (1973), the basement thickness of alluvium is
*6 km near the foothill zone and decreases gradually southwards. Addition to variable
thickness of sedimentary fill, the Ganga Basin exhibits number of ridges, fills and
depressions. The important basement highs are the Delhi-Hardwar ridge in the west, the
Faridabad ridge in the middle, the Monghyr-Ghazipur ridge in the east, a poorly
developed high in the Mirzapur-Ghazipur area and smaller “highs” of Raxaul, Bahraich
and Puranpur. There are two important basins or low areas, namely Gandak and Sarda
depressions (see Fig. 1). The densely populated alluvial plain is characterized by the
presence of several transverse and oblique subsurface faults. A number of subsurface
faults trending NW-SE, WNW-ESE and NE-SW, transverse to the trends of the
Himalayan, were reported and most of these are known to be seismically active
(Valdiya 1987). The IGB is spatially variable in terms of geomorphology, tectonics,
differences in geological deposits, and dynamics of river deposition, all resulting in the
presence of different lithological units. MASW survey has been carried out in the entire
stretch of IGB to capture the spatial variability in shear wave velocity.
Fig. 1. Geological description of Indo Gangetic Basin along with the MASW survey locations
3 Field Survey
MASW is a geophysical survey that records Rayleigh waves on a multichannel record.

It utilizes the dispersion properties of surface waves for determining the shear wave
velocity profile in 1D and 2D format. Based on the source of the surface wave recorder,
MASW is classified as Active or Passive MASW survey. In case of active MASW
survey, surface waves are produced actively through an impact source like a sledge-
hammer. However, in case of Passive MASW survey, surface waves are produced
passively by natural (e.g., thunder and tidal motion) or cultural (e.g., traffic) activities,
(Park et al. 2007). The maximum reliable investigation depth is typically shallower
than 50 m in case of active MASW survey. However, it can range a few hundred
meters in case of passive MASW survey.
The entire process for developing Vs profiles through MASW involves three steps:
procurement of ground roll, development of dispersion curve (a plot of phase velocity
versus frequency), and back calculation (inversion) of Vs profile from the calculated
dispersion curve. Multichannel methods permit a synchronized study of multiple geo-
phone records which allow a transformation from time and space domains to another
domain that acknowledges the recognition of energy peaks including the dispersive
characteristics of the investigating site (Foti et al. 2001). Various approaches are
available for this purpose. The most effective and widely used are the frequency-
wavenumber analysis (f-k) and the spatial autocorrelation method (SPAC) proposed by
Aki (1957). Analogous to f-k analysis, Park et al. (1998) recommended a transformation
algorithm, in which all signals are delayed and summed for a selected phase velocity.
This method helps in the determination of a velocity–frequency diagram that demon-
strates peaks when the expected value coupled the phase velocity of the wave and allows
the identification of dispersion curve. However, this method was generally appropriate
for active measurements. Park and Miller (2008) comprehended this algorithm for
ambient vibration measurement utilizing the linear arrays alongside a roadside. This
method sums up the energy associated with all possible azimuths to construct the
velocity–frequency diagram that enables the abstraction of the dispersion curve.
S-wave velocities fundamentally govern the modification in Rayleigh-wave phase
velocities in case of a layered earth model. Thus, Rayleigh-wave phase velocities can
be used in inverting S wave velocities. Vs profiles are determined from iterative
inversion process which needs the dispersion data and approximations of Poisson’s
ratio and density. Least-squares approach helps in automation of the process (Xia et al.
1999), it further outlines an objective function and a weighting matrix, derives a
solution of minimizing the objective function by the Levenberg–Marquardt (L-M)
method and the singular value decomposition (SVD) technique. An iterative solution of
a weighted damping equation using the L-M method provides a fast and stable solution.
Using the SVD technique, the efficient calculation can be attained by reconstructing a
weighted damping solution. In this method, only Vs is restructured after each iteration,
with Poisson’s ratio, density, however model thickness remaining unaffected
throughout the inversion. As per Xia et al. (1999), the Vs profile is the most important
parameter in the inversion process as fluctuations in density or Poisson ratio give
insignificant effects in dispersion properties.
4 Data Acquisition and Processing
The raw data is acquired for 275 survey sites along the entire stretch of Indo Gangetic
Basin. The data collection covers the major cities of Punjab, Haryana, Uttar Pradesh
and Bihar states of India. Data acquisition sites are shown in Fig. 1. At each location,
both Active and Passive MASW survey has been done to acquire the data at both
shallow and deeper depths. Test setup consists of 24 channel Geode seismographs in
combination with 24 vertical geophones with the frequency of 2.0 Hz. An impulsive
source of 15-pound sledgehammer striking over a 30 cm 30 cm size steel plate
produces surface waves, in case of active survey. Figure 2 shows the acquisition of
active data. However, the amount of energy generated by active source is not sufficient
for producing dispersion curve at lower frequencies (e.g., 5–7 Hz). Hence to get the
investigation depth of about several hundreds of meters, Passive noise is used. Passive
surface waves generated from natural (e.g., tidal motion) or cultural (e.g., traffic) are
used for generating low-frequencies (1–30 Hz). Depending on field conditions and type
of Vs profiles to be obtained (1D or 2D); Passive MASW is divided into two types, i.e.
passive remote and passive roadside MASW surveys. For acquiring the data for Pas-
sive MASW survey for the IGB both passive remote and roadside survey have been
done; based on the availability of place (See Fig. 2). For obtaining the passive data, a
passive roadside acquisition method is used by taking advantage of moving traffic for
producing low-frequency ambient noise. Park and Miller (2008) recommended that
when accomplishing a roadside surface wave using a linear receiver array, a 2-D
dispersion analysis scheme that explains the offline nature of the passive surface waves
is required. Hence, for obtaining the raw data using Passive survey, different sampling
intervals and recording times are used to enhance the dispersion curve quality. After
acquiring the data from both active and passive MASW survey, the individual dis-
persion curves have been extracted using velocity–frequency diagram.
Fig. 2. Multichannel analysis of surface wave instrumentation indicating the sources used in
active and passive survey
The obtained raw data has been further processed to obtain the dispersion curve and
finally to obtain the 1D shear wave velocity profile. Algorithm for obtaining dispersion
curve and Vs profiles through inversion are explained above. For obtaining the Vs
profiles, window-based software named ‘SurfSeis 5’ and ‘ParkSEIS 2’ has been used.
These software process the Rayleigh type seismic wave and generate Vs profiles by
analyzing the fundamental mode of dispersion curve of Rayleigh wave. The dispersion
curve for passive as well as active is given as Fig. 3(a) and (b) respectively. For
(a) enlarging the analyzable frequency (therefore depth) range of dispersion and
(b) identifying the modal nature of dispersion trends, it is useful to combine the
dispersion images of active and passive data. Hence, combining the active and passive
dispersion image has also studied to quantify the depth corresponding to both lower
and upper-frequency range. For most of the data to get the enhanced shear wave
velocity at lower as well as deeper depth combined dispersion is used. The dispersion
curve for combined data is given as Fig. 3(c).
Fig. 3. Typical example of (a) Active, (b) Passive and (c) Combined MASW result for
dispersion and shear wave velocity profile
5 Result and Discussion
5.1 Parametric Study on Dispersion Curve and vs Profiles

These are different ways of acquiring data using MASW. Different parameters, like
recording time, sampling interval, number of geophones, geophone spacing governs the
prediction of dispersion curve. This on other hand affects the layering and shear wave
velocity of the site. Hence, in this study for different sites, these parameters are eval-
uated and variation on dispersion curve and Vs profiles are determined. For recording
the passive data different recording time, sampling intervals, polarity and gain is tested.
For example, data has been acquired for 120 s by changing the sampling interval as
8 ms, 4 ms and 2 ms and gain as high and low gain. Typical dispersion curve and Vs
profiles for four cases have been given as Fig. 4(a). Case A1, A2, A3 and A4 corre-
sponding to recording time interval of 60 s with sampling interval of 10, 8, 4 and 2 ms
respectively. A significant change is observed at lower frequencies, due to that there is
a variation in velocity at deeper depths. Standard deviation of 10 to 25 m/s is observed
within a depth of 50 m. Further for the same site the data is recorded for different time
intervals viz. 10, 30, 60 and 120 s. These cases are respectively referred as Case B1,
B2, B3 and B4 in Fig. 4(b). Significant variation in dispersion curve is observed at
lower frequencies, which affect the velocity at deeper depths. The difference of 150 to
200 m/s is observed at a depth more than 150 m. This may be due to the variability in
traffic density and traffic loading. Hence, while capturing the passive data, recording
must be done for different time intervals to obtain the proper energy bands at all the
frequencies. Similarly, change in gain and for different recording time, there is a change
in dispersion curve and corresponding variation in shear wave velocity. Hence for the
final study, average of all the survey at same location has been used. You can also add
how much error and difference.
Considering the dispersion curves and the corresponding velocity at different layers
for various parameters for data recording following points are concluded
(1) The passive recording must be done for longer period, i.e. more than 60 s with
larger sampling interval
(2) In case the traffic density is low, combined active and passive survey need to be done
(3) In case of poor energy bands, combining two passive images enhances the image to
extract the dispersion curve
(4) To obtain the velocity at shallow as well as deeper depths both active and passive
image need to be combined at the same site
(5) It is always good to do multiple experiments at a site, and average velocity need to
be given.
5.2 Typical Comparison of vs with SPT-N Value and Lithology

SPT is a widely used in-situ test in a borehole to evaluate geotechnical properties of
soil. The preliminary study has been done to compare the SPT N-value with the derived
shear wave velocity at sites in Haryana, Punjab and Uttar Pradesh. All the recorded data
has been used to compare the SPT N value and based on that input setting has been
done for further data recording. The compared SPT N value and recorded Vs profile is
given as Fig. 5(a). For different locations, Vs is compared with the SPT N value data
and typical comparison is presented in Fig. 5(a). It can be seen that recorded shear
wave velocity profiles is following almost the same trend as compared to SPT N value.
The lithology is also compared with the shear wave velocity profile developed in
these regions using MASW and given as Fig. 5(b). On an average six shear wave
velocity profiles are taken near to each of the sites where lithology is known. The
average the shear wave velocity profiles obtained in each region is compared with the
lithology profiles. It can be seen that the for 100 m thick idealized lithology profiles
given by Singh (1996) is compared with the shear wave velocity profiles obtained in
the present study. The Vs profile obtained in this study shows the three-distinct region
Fig. 4. Typical example of variation of dispersion curve and shear wave velocity profile for
variability in (a) sample interval and (b) time of recording data acquisition
(a) a high gradient, low-velocity near-surface region, (2) an intermediate region with
shear wave velocities of about 400–450 m/s, and (3) a high-velocity region at depth
with velocities increasing to 600–800 m/s.
Fig. 5. Typical variation of shear wave velocity with (a) SPT-N value and (b) lithology
5.3 Seismic Site Classification

The shear wave velocity profiles have been developed for 275 sites in Indo Gangetic
Basin. The time-averaged shear wave velocity at top 30 m has been calculated using
PN
di
Vs30 ¼ P 1 ð1Þ
N di
1 Vsi
where, di is the thickness of layer i and Vsi is the shear wave velocity of the layer i and
Vs30 is the time average shear wave velocity in the top 30 m. Time average shear wave
velocity up to 30 m depth has been calculated for all the entire IGB. Figure 6 shows the
variation of Vs30 along the IGB. The sites have been classified as per National Hazard
Reduction Program (NEHRP, BSSC 2003). Profiles with Vs30 [ 1500 m/s,
760\Vs30 1500 m/s, 360\Vs30 760 m/s, 180\Vs30 360 m/s and Vs30 \180 m/s
respectively correspond to Site class A, B, C, D and E.
Fig. 6. Spatial variation of Vs30 along the entire stretch of IGB with seismic site classification as
per NEHRP
About 22% of the sites in Bihar is having Vs30 between 153–180 m/s and classified
as site class E. Majority of the sites in Uttar Pradesh is of site class D, however in
Punjab and Haryana, mostly site class are either C or D. For Punjab and Haryana
region, Vs30 varies from 160 to 180 m/s, 206 to 340 m/s, 382 to 620 m/s and 795 to
1251 m/s respectively for site class E, D, C and B. For Uttar Pradesh, Vs30 varies from
157 to 180 m/s, 247 to 358 m/s, 401 to 630 m/s and 822 to 1136 m/s respectively for
site class E, D, C and B. For Bihar region Vs30 varies from 153 to 180 m/s, 214 to
354 m/s, 387 to 615 m/s and 795 to 1152 m/s respectively for site class E, D, C and B.
Further the spatial variation of Vs30 is compared with the existing geological map of
the IGB (See Fig. 1). Most of the sites near to the Kosi, Satluj and Gomti river is of
seismic site class E; this may be due to the active sedimentation since Quaternary.
Channel alluvium is made up of medium to coarse sand near to Ghaggar river, hence
Vs30 variation is from 180 to 360 m/s. Further near to Sutlej river, low velocity is
observed which is due to the presence of silt and clay along its territory. Most of the
area of Uttar Pradesh having Vs30 between 270 to 450 m/s. It is due to the presence of
Varanasi Older Alluvium of the Quaternary sediments followed disconformably by
Newer alluvium. Most of the sites in the southwest side of Yamuna in UPR is of site
class B, which may be due to the presence of Bundelkhand Faridabad ridge and
Archean granites and Precambrian Metamorphic (See Fig. 1). Similarly, due to the
presence of Munger-Saharsa Ridge and Paleozoic granite in the south of Ganga in
Bihar, most of the sites are either site class B or C. Many of the sites in northeast side of
Uttar Pradesh and Bihar region is of site class B and C, which may be due to the
presence of Siwalik Hills (See Fig. 1).
It can be noticed here that even though Vs30 is widely used for seismic site char-
acterization, but soil depth extends beyond 100 m. So, the applicability of Vs30
moreover, its connection with actual amplification need to be examined for IGB region.
6 Conclusions
In addition to earthquake magnitude and its distance from the epicentre, soil charac-
teristic places a major role in the extent of damages at any site due to any seismic event.
In this study, an attempt has been made to determine the shear velocity profile up to
400 m depth using combined active and passive multichannel analysis of surface wave
survey in 275 selected locations in IGB. Acquisition and processing of the obtained
data has been explained to get the Vs profiles at shallower as well as deeper depths.
Parametric study has been done to identify the variation in dispersion and Vs profiles
by recording time, different sampling interval, different polarity and gain. On an
average the difference of 5 to 50 m/s is observed for shallow depth and 150 to 250 m/s
at deeper depths in Vs It has seen see that changing any parameter during data
recording may result in variation in Vs . Passive recording need to be done for more
than 60 s and to combined active and passive data should be collected in case of low
background noise. Further, these sites are classified and characterized based on time-
averaged Vs in the upper 30 m depth as per NEHRP seismic site classification. For
Punjab and Haryana region, Vs30 varies from 160 to 180 m/s, 206 to 340 m/s, 382 to
620 m/s and 795 to 1251 m/s respectively for site class E, D, C and B. For Uttar
Pradesh, Vs30 varies from 157 to 180 m/s, 247 to 358 m/s, 401 to 630 m/s and 822 to
1136 m/s respectively for site class E, D, C and B. For Bihar region Vs30 varies from
153 to 180 m/s, 214 to 354 m/s, 387 to 615 m/s and 795 to 1152 m/s respectively for
site class E, D, C and B. This is the first time such extensive study was carried out in
IGB for Vs determination at deep and shallow depths.
Acknowledgment. The authors thank the Science and Engineering Research Board (SERB) of
the Department of Science and Technology (DST), India for funding the project titled “Mea-
surement of shear wave velocity at deep soil sites and site response studies”, Ref:
SERB/F/162/2015-2016. Authors also thanked to Geological Society of India for providing the
Geological and Lithological map required for the study. The authors would like to extend their
sincere appreciation to the Deanship of Scientific Research at King Saud University for funding
Research group NO. (RG -1435-09).
References
Aki, K.: Space and time spectra of stationary stochastic waves, with special reference to
microtremors. Bull. Earthq. Res. Inst. 35, 415–456 (1957)
Anbazhagan, P., Sitharam, T.G.: Spatial variability of the weathered and engineering bedrock using
multichannel analysis of surface wave survey. Pure. appl. Geophys. 166(3), 409–428 (2008)
Anbazhagan, P., Kumar, A., Sitharam, T.G.: Amplification factor from intensity map and site
response analysis for the soil sites during 1999 Chamoli earthquake. In: Proceedings of the
3rd Indian Young Geotechnical Engineers Conference, New Delhi, pp. 311–316 (2010)
Bilham, R., Ambraseys, N.: Apparent Himalayan slip deficit from the summation of seismic
moments for Himalayan earthquakes, 1500–2000. Curr. Sci. 88, 1658–1667 (2005)
Bilham, R., et al.: Himalayan seismic hazard. Sci. 293, 1442–1444 (2001)
Boominathan, A., Dodagoudar, G.R., Suganthi, A., Maheshwari, R.U.: Seismic hazard
assessment of Chennai city considering local site effects. J. Earth Syst. Sci. 117(S2), 853–
863 (2008)
BSSC: NEHRP recommended provision for seismic regulation for new buildings and other
structures (FEMA 450). Part 1: provisions, building safety seismic council for the federal
Emergency Management Agency, Washington DC (2003)
Desai, S.S., Choudhury, D.: Spatial variation of probabilistic seismic hazard of Mumbai and
surrounding region. Nat. Hazards 17(1), 1873–1898 (2014)
Desai, S.S., Choudhury, D.: Site-specific seismic ground response study for Nuclear power plants
and ports in Mumbai. Nat Hazards Rev (2015). https://doi.org/10.1061/(ASCE)NH.15276996.
0000177,04015002
Foti, S., Lancellota, R., Socco, L.V., Sambuelli, L.: Application of FK analysis of surface waves
for geotechnical characterization. In: Proceedings of the Fourth International Conference on
Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics and
Symposium in Honour of Professor W.D. Liam Finn, March 26–31, 2001, San Diego,
California, paper no. 1.14, 6 pp (2001)
Humire, F., Sáez, E., Leyton, F., Yañez, G.: Combining active and passive multi-channel analysis
of surface waves to improve reliability of V_S30 estimation using standard equipment. Bull.
Earthq. Eng. 13, 1303–1321 (2015)
Jishnu, R.B., et al.: Ground response analysis of Kanpur soil along Indo-Gangetic Plains (2013).
https://doi.org/10.1016/j.soildyn.2013.04.001
Kumar, A., Anbazhagan, P., Sitharam T.G.: Site-specific ground response study of deep Indo-
Gangetic Basin using representative regional ground motions, Geo-Congress, State of art and
practice in Geotechnical Engineering, Oakland, California, paper no. 1065 (2012)
Kumar, A., Bora, O., Harinarayan, N.H.: Obtaining the surface PGA from site response based on
globally recorded ground motions and matching with the codal values. Nat. Hazards 81, 543–
572 (2016)
Lyon-Caen, H., Molnar, P.: Gravity anomalies, flexure of the Indian Plate, and the structure,
support and evolution of the Himalaya and Ganga Basin. Tecton.S 4, 513–538 (1985)
Pandey, B., et al.: Site characterization of strong motion recording stations of Delhi using joint
inversion of phase velocity dispersion and H/V curve. Bull. Seismol. Soc. Am. (2016). https://
doi.org/10.1785/0120150135
Park, C., Miller, R., Xia, J.: Imaging dispersion curves of surface waves on multi-channel record.
Society of Exploration Geophysicists Expanded Abstracts 1377–1380 (1998)
Park, C.B., Miller, R.D., Xia, J., Ivanov, J.: Multichannel analysis of surface waves (MASW)-
active and passive methods. Lead. Edge 26, 60–64 (2007)
Park, C., Miller, R.: Roadside passive multichannel analysis of surface waves (MASW).
J. Environ. Eng. Geophys. 13, 1–11 (2008)
Rao, M.B.R.: The subsurface geology of the Indo-Gangetic Plains. Geol. Soc. India. 14, 217–242
(1973)
Singh, I.B.: Sedimentological history of Quaternary deposits in Gangetic plain. Indian J. Earth
Sci. 14, 272–282 (1987)
Singh, I.B.: Geological evolution of Ganga plain—an overview. J. Palaeontol. Soc. India 41, 99–
137 (1996)
Valdiya, K.S.: Trans-Himadri thrust and domal upwarps immediately south of collision zone and
tectonic implications. Curr. Sci. 56, 200–209 (1987)
Xia, J., Miller, R.D., Park, C.B.: Estimation of near-surface shear-wave velocity by inversion of
Rayleigh waves. Geophys. 64, 691–700 (1999)
Turbulent Flow Characteristics in Interior
and Wake Region of Emergent and Sparse
Vegetation Patch
Soumen Maji1,2(&), Prashanth Reddy Hanmaiahgari2,

Ram Balachandar3, and Vesselina Roussinova4
1
Department of Civil Engineering, CIT Kokrajhar, Kokrajhar, Assam, India
s.maji@cit.ac.in
2
Department of Civil Engineering, IIT Kharagpur, Kharagpur,
West Bengal, India
3
Department of Civil and Environmental Engineering, University of Windsor,
Windsor, ON N9B 3P4, Canada
4
Department of Mechanical, Automotive and Materials Engineering,
University of Windsor, Windsor, ON N9B 3P4, Canada
Abstract. Vegetation in the flow field affects flow hydrodynamics. For

example, emergent vegetation patch in the flow field produces changes in river
morphology, changes in flow velocity and turbulence. Therefore, investigation
of flow characteristics in and around the emergent vegetation patch is a topic of
importance. The main aim of this study is to investigate the self-similarity and
scaling of profiles of streamwise and lateral velocities, turbulence intensities,
Reynolds shear stress and kinetic energy in the interior and the wake region of
the vegetation patch. A Nortek Vectrinoplus Acoustic Doppler Velocimeter
(ADV) was used to measure the point velocity data throughout the depth along
the centerline of the vegetation patch in the streamwise direction. Seventy
numbers of acrylic cylinders were used to make the emergent vegetation array in
the middle of a laboratory flume. In this study, flows with two different Rey-
nolds numbers and two different aspect ratios were investigated. The results
demonstrate self-similarity of streamwise velocity profiles normalized by depth
averaged velocity in the outer region in the interior and wake region of the
emergent and sparse vegetation patch irrespective of Reynolds number and
aspect ratio of flow. Similarly, wall wake similarity was observed in the Rey-
nolds shear stress distribution.
1 Introduction
Variation of turbulent flow characteristics along the streamwise centerline of the

vegetation patch in the interior and wake region is investigated for two Reynolds
numbers and two aspect ratios on a rough bed. Significance of this study is to inves-
tigate the similarity of turbulence characteristics in the interior of the vegetation at
various flow conditions. Further, scaling of turbulence characteristics throughout the
depth is investigated. The non-dimensional similarity profiles can be used in devel-
oping analytical solutions to similar vegetated flows.

https://doi.org/10.1007/978-3-030-02032-3_7
74 S. Maji et al.
Fox and Stewart (2014) carried out mixed scaling and similarity analysis for tur-
bulent open-channel flow over gravel and cobble beds using the asymptotic invariance
principle (AIP). Cheng et al. (2012) investigated scaling of velocity profile and velocity
defect in sparse submerged rigid vegetation canopy, in which vertical height is scaled
by height of the vegetation, and proposed a power law for the velocity profile with
exponent limited between 2.5 and 9. Tang and Hsieh (2015) developed analytical
solutions for the velocity profiles, shear stress profiles and kinetic energy dissipation
rates in submerged vegetation as well as emergent vegetation low Reynolds number
flows using non-dimensional governing equations. Maji et al. (2016, 2017) studied
channel flow with emergent vegetation patch, mainly focusing on streamwise velocity
variation along the transverse direction. Meftah and Mossa (2016a, b) proposed a log
law for the depth averaged streamwise velocity in the shear layer adjacent to the patch
as a function of transverse distance from the edge of the patch. However, Meftah and
Mossa (2013, 2016a, b) study is limited to the transversal flow velocity profile influ-
enced by momentum transfer in the form of shear stress at the interface between the
array and the unobstructed flow. Kubrak et al. (2008) developed a one dimensional
model to study the velocity profile for submerged and emergent vegetation using the
mixing length. Huai et al. (2009a, b) proposed a numerical model for submerged and
emergent rigid vegetation using von Kármán similarity.
In this study, special attention is given to the similarity of velocity and turbulence
quantities in the interior of a sparse and emergent vegetation patch which is lacking in
the literature. The content of this study is arranged as follows. Experimental setup and
methodology of experimental data collection at various streamwise locations in the
upstream, interior and downstream of the sparse and emergent vegetation patch to
investigate hydrodynamics is explained in experimental methodology section. In the
results and discussion, similarity of velocity profiles, turbulence intensities, Reynolds
shear stress have been investigated along the streamwise centerline of vegetation patch
and the wake region. Finally, results and discussion are summarized in conclusions
section.
2 Experimental Methodology
All experiments were performed in a wide rectangular channel (12 m long, 0.91 m
wide and 0.61 m deep) existing in the Hydraulic and Water Resources Engineering
Laboratory of Indian Institute of Technology Kharagpur, India. The flume contained
two glass sidewalls (8 m long and 15 mm thickness) for the clear view of the flow. The
test section of the flow was considered as 3 m long which begins from 7 m down-
stream of the inlet. The water depths (h) in the different experiments were maintained
as 150 mm and 200 mm by a tailgate which was fixed at the downstream end of the
flume. The maintained average flow velocity was 0.29 m/s and 0.31 m/s respectively
for different experiments. The cross-sectional lines through those streamwise locations
were located in the upstream (x = 0 m, 0.50 m, 0.90 m, 1.00 m), interior (x = 1.30 m,
1.50 m, 1.78 m, 1.90 m) and downstream (x = 2.20 m, 2.50 m, 2.85 m and 3.00 m) of
the vegetation patch. Figure 1a shows the photograph of vegetation patch placed in the
Turbulent Flow Characteristics in Interior and Wake Region 75
middle of the flume. Array of acrylic circular rods was used to simulate rigid and
emergent vegetation patch. To prevent the vibrations, the acrylic rods top ends were
properly fixed. Table 1 gives the complete details of the vegetation patch.
Fig. 1. (a) Photograph of vegetation patch and Acoustic Doppler Velocimetry (flow direction is
from bottom to top). (b) Top view schematic of streamwise measurement locations
In this study, turbulence along the centerline of the vegetation patch which is
0.474 m from the RHS flume wall (RHS is with reference to the looking in the flow
direction) has been investigated (refer Fig. 1b). Refer to Fig. 1b for (not to scale) top
view of the streamwise locations where experimental data were measured. Velocities
76 S. Maji et al.
Table 1. Details of vegetation patch

Vegetation length (x) 0.816 m
Vegetation width (y) 0.246 mm
Rods per unit area 350 rods/m2
Length of vegetation cylinder 300 mm
Diameter of vegetation cylinder 6.4 mm
Solid volume fraction 0.00894
were measured with the Acoustic Doppler Velocimetry. Flow depths were determined
by a graduated millimeter scale with an accuracy ±0.1 mm.
Velocity was measured at fifteen vertical locations which were z = 3, 5, 7, 9, 15,
20, 25, 30, 40, 50, 60, 70, 80, 90 and 100 mm for water depth 150 mm (in addition
120 mm and 150 mm for water depth 200 mm) from the channel bed. Data was
recorded at each vertical location for 300 s with a sampling frequency of 100 Hz and
sampling volume of 2.5 103 mm3. From the velocity data recorded, random spikes
were removed by the spike filtering technique as suggested by the previous researchers
(Goring and Nikora 2002; Wahl 2003; Mori et al. 2007). The removed data was
replaced by the values obtained from cubic interpolation. To ensure the quality of the
data obtained after spike filtering, the signal-to-noise ratio test as proposed by Chanson
et al. (2008) was implemented. The signal-to-noise ratio was maintained as 15, and the
data with correlation values greater than 75% were considered.
3 Results and Discussion
Three experiments with varying Reynolds number (Re) and aspect ratio (Ar) were
conducted. The experiments are denoted as Re1Ar1, Re1Ar2 and Re2Ar2. All
experimental conditions are given in Table 2. Experimental data, including interior and
wake region were plotted together to study similarity of important flow variables and
other parameters of the flow field for the varying Reynolds number and aspect ratio.
Table 2. Details of flow conditions in the laboratory experiments

Exp. h (mm) u (m/s) umax (m/s) u (m/s) Mean flow Re Fr Ar
Re1Ar1 155 0.29 0.30 0.238 43500 0.235 5.87
Re1Ar2 200 0.22 0.25 0.197 43500 0.157 4.55
Re2Ar2 200 0.31 0.34 0.254 62000 0.221 4.55
3.1 Similarity of Velocity Profiles

Distributions of streamwise (u) and lateral velocities (v) are scaled with depth averaged
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
velocity ðuÞ and velocity vector v2 þ w2 in the lateral direction respectively to
study the self-similarity of velocities in the interior and the wake region of the
vegetation patch. Henceforth, the scaled streamwise velocity profiles are plotted in
Fig. 2a and b for interior and wake regions respectively. The streamwise velocity
profile is uniform and rectangular in shape in the interior region (Fig. 2a), whereas
velocity profile follows a power law in the wake region (Fig. 2b). In the interior of the
vegetation patch similarity of the streamwise velocity profile is observed in the outer
layer but not in the inner layer (Fig. 2a). For a given Reynolds number, with an
increase in aspect ratio, velocity profile is attaining a uniform velocity throughout the
depth within a short distance from the leading edge. Similarly, for a given aspect ratio,
with an increase in Reynolds number, velocity profile becomes uniform throughout the
depth within a short distance from the leading edge. Thus, there is a disagreement of
velocity profiles in the inner layer at different streamwise locations in the interior of the
vegetation patch.
In the wake region of vegetation patch, similarity of the streamwise velocity profile
is observed in the outer layer but not in the inner layer because velocity profile is trying
to attain the fully developed velocity profile from the rectangular velocity profile as the
distance increases in the downstream direction. For a given Reynolds number, with an
increase in aspect ratio, the distance required for rectangular velocity profile to attain
fully developed velocity profile is becoming shorter. Similarly, for a given aspect ratio,
with an increase in Reynolds number, the downstream distance required for rectangular
velocity profile to attain fully developed is decreasing. Thus, there is a disagreement of
velocity profiles in the inner layer at different streamwise locations in the wake of the
vegetation patch.
It is observed from Fig. 2c and d that lateral velocity vector in the wake region
follows the same trend as in the interior of the vegetation. The scatter in the data is also
similar in the interior and wake regions. The scaled lateral velocity in the interior of the
vegetation is increasing in the streamwise direction, whereas the scaled lateral velocity
in the wake region is decreasing with increase in the streamwise distance. In addition,
the scaled lateral velocity is increasing with increase in the aspect ratio. The probable
reason behind this characteristic is that the intensity of lateral flow increases along the
centerline of the vegetation patch as the flow increasingly deviates from the central
region towards the sidewalls, owing to the resistance offered by the vegetation patch,
however, the strength of the lateral flow is decreasing in the wake region along the
downstream direction because of diminishing wake effect along the streamwise
direction.
3.2 Turbulence Intensities

Upstream of the vegetation patch, the flow along the centerline is irrotational and slows
down due to the blocking effect of the entire patch and the drag force from the
cylinders. The flow bleeds through the vegetation with low velocity, and high-velocity
flow passes through the region without vegetation i.e., the region between sidewalls
and vegetation patch. Distributions of turbulence intensities (u0 , v0 and w0 ) are scaled
pffiffiffi
with RMS value of kinetic energy k in the interior and wake region for all the
experimental runs have been shown in Fig. 3(a, b, c, d, e and f). Observing the Fig. 3, it
has been concluded that distributions of turbulence intensities in the interior and wake
78 S. Maji et al.
(a) (b)
(c) (d)
(Re1Ar1, 1.3) (Re1Ar1, 1.5) (Re1Ar1, 1.78) (Re1Ar1, 1.9)

(Re1Ar2, 1.3) (Re1Ar2, 1.5) (Re1Ar2, 1.78) (Re1Ar2, 1.9) Interior
(Re2Ar2, 1.3) (Re2Ar2, 1.5) (Re2Ar2, 1.78) (Re2Ar2, 1.9) Region
(Re1Ar2, 2.2) (Re1Ar2, 2.5) (Re1Ar2, 2.85) (Re1Ar2, 3.0) Wake
Fig. 2. Streamwise velocity distribution profiles throughout depth (a) interior and (b) wake
region, and lateral velocity vectors throughout depth (c) interior and (d) wake region
region along the centerline of the patch in the streamwise direction follow the same
geometry. Change in Reynolds number and aspect ratio found to be affecting the
turbulence intensity profiles. In Fig. 3a, streamwise turbulence intensity is decreasing
in the streamwise direction. Moreover, the streamwise turbulence intensity in the
interior of the vegetation found to be increasing with increasing aspect ratio and
Reynolds number. Streamwise turbulence intensity variation in the wake region is
shown in Fig. 3b. In the wake region streamwise turbulence intensity is increasing in
the downstream direction. Similar to the interior region, the streamwise turbulence
intensity in the wake region is found to be increasing with increasing aspect ratio and
Reynolds number. Streamwise turbulence intensity is the maximum near the bed and
decreasing towards the free surface in the interior and the wake region. In contrast,
lateral and vertical turbulence intensities are increasing with increasing vertical coor-
dinate. The lateral turbulence intensity in the interior of the vegetation is increasing in
the downstream direction which is contrary to the behavior of streamwise turbulence
intensity (Fig. 3c). The lateral turbulence intensity in the interior of the vegetation
patch is increasing with increasing aspect ratio. The lateral turbulence intensity profile
in the wake region is similar to that in the interior of the vegetation patch. The lateral
turbulence intensity in the wake region is decreasing with increase in the streamwise
distance. Similar to the behavior in the interior of the patch, the lateral turbulence
intensity in the wake region is increasing with increasing aspect ratio. Vertical tur-
bulence intensity profiles in the interior and wake regions are shown in Fig. 3e and f.
Vertical turbulence intensity inside the vegetation patch and in the wake region is
independent of the streamwise location, however, w0 is increasing with increase in
Reynolds number.
3.3 Reynolds Shear Stress
Reynolds shear stress u0 w0 plays a dominant role in and around the vegetation patches
compared to other two Reynolds shear stresses i.e. u0 v0 and v0 w0 . Therefore, vertical
distributions of Reynolds shear stress u0 w0 along the midsection of the channel for
interior of the vegetation and wake region have been shown in Fig. 4(a) and (b). In

these plots, maximum Reynolds shear stress u0 w0 max is used for scaling the
Reynolds shear stress. The self-similarity of the RSS profile is poor in the interior of the
vegetation patch as shown in Fig. 4a. The magnitude of u0 w0 in the interior vegetation
(Fig. 4a) is more than the wake region (Refer Fig. 4b) due to intense turbulent mixing
and high interaction between the vortices in the interior of the vegetation patch. In the
interior of vegetation patch, Reynolds shear stress u0 w0 decreases as the distance
increases in the downstream and this trend continues up to the trailing edge of the
vegetation patch. In the interior of the vegetation, the Reynolds shear stress u0 w0 is
found to be increasing with increasing aspect ratio and Reynolds number. In the interior
of the vegetation patch, peak Reynolds shear stress occurring away from the bed which
shows higher thickness of roughness sublayer in the interior.
In the wake region, the Reynolds shear stress profile follows an exponential curve
which decreases as the vertical coordinate increases as shown in Fig. 4b. The self-
similarity of the Reynolds shear stress in the wake region is better than that in the
interior of the vegetation. In the wake region of vegetation patch, Reynolds shear stress
u0 w0 increases as the distance increases in the downstream as the flow is recovering
from the wake created by the porous vegetation patch. In the wake region, the peak of
RSS occurs close to the bed which indicates thin roughness sublayer.
80 S. Maji et al.
(a) (b)
(c) (d)
(e) (f)

Fig. 3. Streamwise turbulence intensity profile along depth in (a) interior vegetation patch and
(b) wake region; lateral turbulence intensity profile in (c) interior and (d) wake region; and
vertical turbulence intensity in (e) interior and (f) wake region
(a) (b)

Fig. 4. Vertical distribution of Reynolds shear stress in (a) interior and (b) wake region (Thick
line shows theoretical linear distribution in uniform open channel flow)
4 Conclusions
Similarity of streamwise velocity profile scaled by depth average velocity is observed

in the outer layer in the interior and wake of the vegetation patch irrespective of aspect
ratio and Reynolds number. The similarity of the lateral velocity profile scaled by the
lateral velocity vector is not very good. The scaled lateral velocity in the interior of the
vegetation is increasing in the streamwise direction, whereas the scaled lateral velocity
in the wake region is decreasing with increasing streamwise distance.
Self-similarity is observed in the turbulence intensities scaled by kinetic energy.
The streamwise, lateral and vertical turbulence intensities in the interior and wake
region are found to be increasing with increase in aspect ratio and Reynolds number.
Reynolds shear stress is normalized by shear velocity. The self-similarity of the
Reynolds shear stress in the wake region is better than that in the interior of the
vegetation. In the wake region, the Reynolds shear stress profile follows an exponential
curve with maximum near the bed. Reynolds shear stress is found to be increasing with
increase in the Reynolds number.
References
Chanson, H., Trevethan, M., Aoki, S.I.: Acoustic Doppler Velocimetry (ADV) in small estuary:
field experience and signal post-processing. Flow Meas. Instrum. (2008). https://doi.org/10.
1016/j.flowmeasinst.2008.03.003
Cheng, N.-S., Nguyen, H.T., Tan, S.K., Shao, S.: Scaling of velocity profiles for depth-limited
open channel flows over simulated rigid vegetation. J. Hydraul. Eng. (2012). https://doi.org/
10.1061/(ASCE)HY.1943-7900.0000562
82 S. Maji et al.
Fox, J.F., Stewart, R.L.: Mixed scaling for open-channel flow over gravel and cobbles. J. Eng.
Mech. (2014). https://doi.org/10.1061/(ASCE)EM.1943-7889.0000793
Goring, D.G., Nikora, V.I.: Despiking acoustic Doppler velocimeter data. J. Hydraul. Eng.
128(1), 117–126 (2002). https://doi.org/10.1061/(asce)0733-9429(2002)128:1(117)
Huai, W.X., Chen, Z.B., Jie, H.A.N., Zhang, L.X., Zeng, Y.H.: Mathematical model for the flow
with submerged and emerged rigid vegetation. J. Hydrodyn. (2009a). https://doi.org/10.1016/
s1001-6058(08)60205-x
Huai, W.X., Zeng, Y.H., Xu, Z.G., Yang, Z.H.: Three-layer model for vertical velocity
distribution in open channel flow with submerged rigid vegetation. Adv. Water Resour.
Elsevier (2009b). https://doi.org/10.1016/j.advwatres.2008.11.014
Kubrak, E., Kubrak, J., Rowiński, P.M.: Vertical velocity distributions through and above
submerged, flexible vegetation. Hydrol. Sci. J. 53(4), 905–920 (2008). https://doi.org/10.
1623/hysj.53.4.905
Maji, S., Pal, D., Hanmaiahgari, P.R., Pu, J.H.: Phenomenological features of turbulent
hydrodynamics in sparsely vegetated open channel flow. J. Appl. Fluid Mech. (2016). ISSN
1735-3572, EISSN 1735-3645
Maji, S., Pal, D., Hanmaiahgari, P.R., Gupta, U.P.: Hydrodynamics and turbulence in emergent
and sparsely vegetated open channel flow. Environ. Fluid Mech. 17, 853–877 (2017). https://
doi.org/10.1007/s10652-017-9531-2
Meftah, M.B., Mossa, M.: Prediction of channel flow characteristics through square arrays of
emergent cylinders. Phys. Fluids (2013). https://doi.org/10.1063/1.4802047
Meftah, M.B., Mossa, M.: A modified log-law of flow velocity distribution in partly obstructed
open channels. Environ. Fluid Mech. (2016a). https://doi.org/10.1007/s10652-015-9439-7
Meftah, M.B., Mossa, M.: Partially obstructed channel: contraction ratio effect on the flow
hydrodynamic structure and prediction of the transversal mean velocity profile. J. Hydrol.
542, 87–100. (2016b). https://doi.org/10.1016/j.jhydrol.2016.08.057
Mori, N., Suzuki, T., Kakuno, S.: Noise of acoustic Doppler velocimeter data in bubbly flows.
J. Eng. Mech. (2007). https://doi.org/10.1061/(asce)0733-9399(2007)133:1(122)
Tang, C.Y., Hsieh, P.C.: Dynamic analysis of vegetated water flows. J. Hydrol. Eng. (2015).
https://doi.org/10.1061/(ASCE)HE.1943-5584.0001294
Wahl, T.L.: Discussion of “Despiking acoustic doppler velocimeter data” by Derek G. Goring
and Vladimir I. Nikora. J. Hydraul. Eng. (2003). https://doi.org/10.1061/(asce)0733-9429
(2003)129:6(484)
Volcanic Disaster and the Decline of Mataram
Kingdom in the Central Java, Indonesia
Sari Bahagiarti Kusumayudha(&), Helmy Murwanto, Sutarto,

and Siti Umiyatun Choiriyah
Universitas Pembangunan Nasional Veteran Yogyakarta, Yogyakarta, Indonesia

saribk@upnyk.ac.id
Abstract. In the area of Central Java and Yogyakarta there are several volca-
noes namely Sindoro volcano, Sumbing volcano, Slamet volcano, Ungaran
volcano, Merbabu volcano, and Merapi volcano. Among the volcanoes, Merapi
volcano is the most active, generating lava flows, pyroclastic flows, glowing
clouds, and lahar. On the other hand, during the 7th to 10th century in this
region had been existed an ancient monarchy, known as the Mataram Kingdom.
Its glory at that time was marked by many ancient heritages especially in the
form of temples. There are many monumental, beautiful, and majestic archae-
ological relics, located on the plains, slopes, even near the summit of Merapi
volcano, Sumbing volcano, Sindoro volcano, Dieng mountains, and Ungaran
volcano.
In the 11th century the history of Mataram Kingdom was not recorded any-
more, suspected that the kingdom declined, and the cultural center of excellence
transferred to East Java. Up to now, the cause of the collapse of Mataram
Kingdom is still a mystery. Some historians suspect as a result of a great war,
and some others thought as the impact of volcanic disaster. Nevertheless, the fall
of Mataram Kingdom still being an enigma.
At the southern slopes of Merapi volcano, there are some ancient buildings
that buried by volcanic deposits. For example Kadisoka temple part of the
building is still dotted sandstone, tuff, and pyroclastic sandstone, cobbles.
Similarly, Sambisari, Kedulan, and Pustakasala temples, when discovered, they
were sinked by sand, tuff, and fluvio-volcanic sediments. On the slopes of
Sindoro volcano, there are also ancient sites, namely Liyangan was covered by
pyroclastic flow, tuff, and lahar deposits. Thus it is possible that volcanic dis-
asters have contributed to the decline of Mataram Kingdom.
Keywords: Volcanic disaster Pyroclastic flow Lahar Ancient temples

Mataram Kingdom
1 Introduction
In the Central Java and Yogyakarta Special Territory of Indonesia, during the 7th to
10th century, there was an independent state, namely the Mataram Kingdom. The past
kingdom glory is marked by the existing of many ancient relics and buildings, two of
which are internationally well-known, as Borobudur temple and Prambanan temple.
Some beautiful and magnificent monumental archaeological remains have been found

https://doi.org/10.1007/978-3-030-02032-3_8
84 S. B. Kusumayudha et al.
in the study area. The beauty and splendor of the ancient heritage illustrates the taste of
architectural artworks of noble, in the form of Buddhist and Shiwa Hinduist temples.
The temples are generally scattered on the plains, the feet slopes, the central slopes, the
upper slopes, and some even on the summit of the mountains. At the time of its
discovery, several temples were heavily damaged or buried by soil and rocks as deep as
2 to 6 m.
There were no archaeological relic discoveries after the 10th century, indicating
that in the Central Java did not exist any governmental or state. This was a marker of
the sudden disappearance of the ancient Mataram Kingdom from historical records. It
was subsequently replaced by the emergence of various kingdoms in the eastern Java,
starting from the kingdoms of Singosari, Kediri, Jenggala, to Majapahit, from the 12th
century to the 14th century. The kingdom history seemed to be allegedly migrated to
the East Java.
According to historians, the movement of the central government from the Central
to the East Java was driven by political, cultural and religious aspects. In the matter of
facts, some ancient relics were found in a collapsed condition, or buried by sediments
of natural works. Therefore, in the framework of providing historical corrections and
information, geological-based studies was done to unravel the enigma of the Mataram
decline and its relationship with natural disaster in the Central Java and Yogyakarta
Special Territory.
Objectives of this study are to mapping and analyzing the geological disaster tracks
and records such as volcanic eruption, lava, and ash deposits around ancient archae-
ological sites that suspected of contributing to the damage the temples. The location of
the study area is shown in Fig. 1.
2 Literature Reviews
2.1 History of the Mataram Kingdom

The Mataram Kingdom in the Central Java, established from the 7th to the 10th
century, was ruled by kings from the Sanjaya dynasty and the Syailendra dynasty. It is
written in the Mantyasih inscription (907 AD), that the Ancient Mataram Kingdom
authority was covering Central Java and Yogyakarta territories. But in the late 10th
century the royal capital moved to the East Java, and the cultural center of Central Java
seemed to shift to East Java, marked by the emergence of Medang, Kahuripan, Kediri,
Jenggala, Singosari, and Majapahit kingdoms. The history then returned to the Central
Java after Majapahit Kingdom collapsed, entering the Islamic era, starting from Demak
Kingdom, Pajang Kingdom, and Islamic Mataram Kingdom founded by Panembahan
Senopati (Kusumayudha 2006a, b).
Poesponegoro and Notosusanto (2008) stated that at the time of the Mataram
Kingdom under the king Rakai Sumba Dyah Wawa which ruled in 919–925 AD
(inscription Air Kali, inscription Kinawe, 928 AD, inscription Kambang Sri Jebung,
and inscription Biota), the kingdom undergone a great natural event. The continuity of
the government ended suddenly. This is allegedly due to the eruption of Merapi vol-
cano resulting in earthquakes and lahar flooding. The event is expected to result in the
Volcanic Disaster and the Decline of Mataram Kingdom 85
Fig. 1. Map showing the location of the study area
royal center being moved to East Java. In another inscription it is written that the
transfer of government occurred in 930 AD, during the reign of the last king of the
Mataram Kingdom, Mpu Sindok (928–947 AD). Mpu Sindok subsequently established
Medang Kingdom where the center of his government was initially located in the
upstream of River Brantas (Poesponegoro and Notosusanto 2008).
According to Van Bemmelen (1949), Neuman van Padang (1951) in Kusumayudha
(2006a, b) and MacDonald (1972), Merapi volcano erupted catastrophically in 1006.
This eruption resulted in a total paralysis of the economy, farmland severely damaged,
and volcanic deposits 6 meters thick covering the ground. It was illustrated that the
people and the royal family were destroyed by the lahar floods and buried by volcanic
ash.
Nevertheless, this hypothesis was opposed by some volcanologists. Based on
inscription Pucangan, during the reign of Dharmawangsa (1016–1017 AD), the king-
dom suffered a major catastrophe causing the capital to be destroyed, many deaths, and
economical paralyzed. In the event, the cousin of the king, named Airlangga was able
to escape to the Southern Mountains. Airlangga then develop the Singosari Kingdom.
This happening is known as the Maha Pralaya (Zen 2006). With regard to the greatest
of praws, related to the decline of the Mataram Kingdom, there are still two different
opinions, whether due enemy attacks from the Wura-wari Kingdom (Sri Wijaya) or by
the eruption of Merapi volcano (Kusumayudha 2006a, b).
2.2 Volcanic Disaster

Kusumadinata (1979) recorded a series of Merapi eruptions proving that the volcano has
been very often erupted. The volcanic activity rest time is usually 5 years. During the
eruption break, Merapi will slowly build a lava dome that will be destroyed and ruptured
in the next eruption period. In the valley of River Sileng, Borobudur area, there is found
such volcanic ash as thick as 2 m. From the results of radio-dating testing, the sediment
was estimated to be approximately 600–700 years old. Some evidences show that at that
time there was such a major eruption of Merapi volcano. Merapi eruptions which
claimed many lives, recorded in history are as follows (Table 1):
Table 1. The records of Merapi eruptions causing lost (Kusumayudha 2013a, 2013b)
Year Eruption characteristics Number of
life lost
1672 Producing pyroclastic flows, glowing clouds, and lahar 300
1930–1931 Normal eruptions, lava flows, pyroclastic flows, glowing clouds, and lahar 1369
1954 Producing pyroclastic flows, glowing clouds, ash and lapilli falls 64
1961 Producing lava flows and glowing clouds 6
1969 Explosive eruption, glowing cloud of explosion, lava dome collapse, bomb 3
and block falls, and ash rain
1972–1973 Black smokes 3 km high above the summit of the volcano, sand and stones
falls at the Babadan observatory, lava dome collapse producing pyroclastic
flow, glowing clouds to River Batang reaching 3 km distant
1994 Eruption resulting pyroclastic flows and glowing clouds to the south from the 67
summit to River Boyong, reaching 6 km distance
2006 Merapi type eruption, producing pyroclastic flows of 4 km distance to the 2
valley of River Gendol, glowing clouds, and lahar
2010 Pelean type eruption, producing lava flows, pyroclastic flows of 17 km 250
distance to the valley of River Gendol, glowing clouds, lapilli and ash falls,
lahar
The prehistoric major eruptions of Merapi were never known for certain by anyone.
Scientists can only try to trace and reconstruct these occurrences with geological, vol-
canological, historical, archaeological, cultural, even metaphysical approaches. For
human life, actually the more important thing is not about when Merapi erupted exactly,
but rather how the impact of the eruption to human life (Kusumayudha 2006a, b).
Borobudur temple has a different historical background. Murwanto et al.
(2004) found evidences related to the existence of a lake around the temple at that time.
According to Van Bemmelen (1949), the ancient Borobudur Lake occurred as a result
of a major eruption of Merapi volcano in 1006 AD. This devastating eruption caused
apart of the peak of Merapi to collapse to the southwest direction, stemming the flow of
River Progo, forming a large lake in the area of South Kedu. The Borobudur great lake
as a hollow between the mountains (intermontains basin) was gradually deterred by
tectonic and sedimentation processes. In the VIII century. Borobudur temple was built
on a small hill surrounded by a lake during the reign of King Smaratungga of the
Syailendra dynasty.
The activity of Merapi volcano has a major contribution to the environmental

change of the lake and the mainland. The eruption materials gradually hoarded the
Borobudur Lake, causing it to become shallow and finally dry at the end of the XIII
century. Merapi eruption materials not only dried the lake, but also damaged and buried
Borobudur temple as those of other temples in the Central Java and Yogyakarta.
In addition to the primary hazards of pyroclastic flows, glowing clouds and vol-
canic ash, the secondary threat of volcano that is no less dangerous is lahar. Lahar is a
dense stream that occurs when the pyroclastic deposits in the upper slopes of the
volcano mixes with rain water becoming saturated, then influenced by gravity to
flowing downstream through the river valleys as a heavy currents of mud and rocks
(Kusumayudha 2013a, b). Lahar behaves enormous erosion, high destructiveness,
capable of transporting very large materials, including boulders, trees, and buildings.
Lahar of Merapi in 1930–1931 along with the primary eruption, has claimed of 1369
deaths (Kusumadinata 1979)
3 Methods of Study
The research carried out with descriptive, comparative, and surveys methods that was
complemented by geological phenomena mapping. The data used consist of secondary
data and primary data. Secondary data includes various information from the existing
studies. Primary data obtained through survey and field mapping. Geological studies to
be carried out include geomorphology, petrology, stratigraphy, sedimentology, geo-
logical structure, and disaster analyses.
The stages to be carried out during the research were divided into three steps.
Firstly, to collect secondary data consist of map showing the distribution of ancient
relics, regional geological map, tectonic map, disaster map, aerial photography/satellite
imagery, and historical research data ever undertaken by other researchers. Secondly, to
analyse the geological map, tectonic map, and disaster map as well as remote sensing
interpretation using aerial photography/satellite imagery against the spread of archae-
ological sites, and the presence of geological structures around the sites. Thirdly, to
mapping geological details around the archaeological sites, including the dissemination
of sediment produced by volcanic activity, either the primary activity of pyroclastic or
the secondary activity of lahar.
4.1 Archeological Sites Discoveries

Some monumental relics sites of the ancient Mataram Kingdom were formerly found in a
collapse conditions or buried by soil. Remnants of the Mataram Kingdom glory can be
traced from the discovery of many ancient relics in the Central Java and Yogyakarta
(Fig. 2).
Fig. 2. Distribution of temples and archaeological sites of Java (www.wikipedia.org/wiki/

Daftar_candi_di_Indonesia)
4.1.1 Temples on the Foot Slopes of Merapi

In the year 1966 Sambisari temple was found by a farmer. When discovered, it was
buried by sediments of about 6 m thick. The sediments covering the temple comprise
volcanic materials including sand, tuff, and stones with the grainsize of granule to
boulder. As it is displayed on the bank of River Kuning, the deposits show various
sedimentary structures such as imbrication and carpet tract, indicating that they are
lahar deposit.
At another village namely Kadisoka, there is a small Hindu temple that most of the
building still hidden by eruption deposits of Merapi volcano (Fig. 3). In the
embankment there are sedimentary structures which indicate that these materials are the
interlayering of lahar, fluvial, pyroclastic falls, and glowing clouds deposits. According
to Mulyaningsih (2006a, b), around the Kadisoka Temple, there is an evidence of 4
(four) sequences of Merapi deposit that buried the temple, each of which occurred in
the different periods. The oldest eruption period happened about 1900 years ago, or in
the II century.
Fig. 3. Kadisoka temple (left) and Pustakasala temple (right) (Doc: the authors)
There is another temple called the Kedulan temple, has founded in the IX century.
As that of Sambisari temple, it was covered by about 6 meters thick of volcanic
materials. It consists of sand, granule, gravel, and boulders with tuff impurities. Based
on the soil profile analyzed by Mulyaningsih (2006a, b), the deposits display an
interlayering of lahar and pyroclastic fall sedimentation. In a different time and place,
when the Islamic University of Indonesia (UII) would built a library building, there also
was found a hidden temple which then given the name Pustakasala (Fig. 3). The ruins
of the ancient temple were covered by laharic and pyroclastic deposits of Merapi
volcano as well.
4.1.2 Sites on the Foot Slope of Sindoro Volcano

On the southwestern slopes of Sindoro volcano, Liyangan site was found by the sand
miners (Fig. 4). This site was buried by volcanic deposits as thick as more than 10 m.
Based on the findings of wood charcoal, the ash content at the sediment that buried the
site, it can be concluded that the Liyangan Site is buried by eruption products of
pyroclastic falls and pyroclastic flows.
Fig. 4. Liyangan archaeological site, on the flank of Sindoro volcano (Doc: the authors)
4.2 Geology and Volcanism in the Central Java

The Java island tectonic setting is controlled by the subduction of the India – Australia
plate beneath the Eurasia plate forming a series of active volcanoes stand in the middle
part of the island (Fig. 5). Despite there are plenty of active volcanoes, due to this
tectonic setting, the Java Island is also subjected to earthquake very often. At the
present time, the central Java area still affected by several geological disasters such as
2006 Yogyakarta earthquake, and 2010 Merapi eruption. In the study area, presently,
volcano that still active is only Merapi, while Merbabu volcano, Sumbing volcano and
Sindoro volcano are classified to be in post volcanism stage.
Merapi volcano, is one of the most active volcanoes in Indonesia. When it erupts,
the volcano occasionally spews lava, glowing clouds, and pyroclastic materials out.
Merapi always offers opportunities as well as threats to the life of the surrounding
community. It is able to act as the source of prosperity and the source of the disaster as
well. The magma chamber of Merapi is very shallow, it is supplied continuously by the
partial melting of the India-Australia ocean plate that constantly moves to the North,
Fig. 5. The tectonic setting of the Java Island, in the convergence of India-Australia plate and
Eurasia plate (www.earthobservatory.sg)
subducted under the of the Eurasia continental plate, making the volcano remains alive
(Kusumayudha 2013a, b). The distribution of Merapi eruptions deposits are displayed
in the geological map of the volcano, as shown in Fig. 6, while situation and condition
related to Merapi activities are presented in Figs. 7, 8, 9, and 10.
Fig. 6. The geological map of Merapi Volcano and the surrounding area (After vide
Kusumayudha et al. 2009)
Fig. 7. Satellite image showing Merapi eruption and lava flow. Acquisition date: November 11,
2010 (courtesy: DigitalGlobe)
Fig. 8. Lahar deposits of 1994 (left) and 2006 (right) (Doc: the authors)
Fig. 9. Kaliadem village buried by pyroclastic deposits, 2010 (Doc: the authors)
Fig. 10. After the eruption of 2010 (left), Borobudur temple covered by volcanic ash (right)
(Doc: the Authors)
5 Conclusions
Based on the above discussion, it can be concluded as follows:

1. Some temples in the Central Java and Yogyakarta were discovered in damage
condition and/or buried by volcanic materials. They are Sambisari temple, Kadisoka
temple, Kedulan temple, Pustakasala temple, and the archeological site of Liyangan.
2. Merapi volcano is one of the most active volcanoes of Indonesia, its eruption in
2006 and 2010 produced pyroclastic flows with a range of about 8 to 17 km
distance, burying Kaliadem village of Cangkringan district and surrounding areas,
and resulted in areas with a radius of 30 km covered by volcanic ash.
3. Referring to the data that many temples were buried by volcanic materials, it can be
concluded that eruptions of volcanoes, especially Merapi volcano had taken part in
the destruction of the glory of Mataram Kingdom in the Central Java and
Yogyakarta.
References
Kusumadinata, K.: Data Dasar Gunungapi Indonesia, Departemen Pertambangan dan Energi R.I
(1979)
Kusumayudha, S.B.: Merapi: A Beautiful Bounty. Penerbit PT Citra Adi Parama, Yogyakarta
(2013a)
Kusumayudha, S.B.: Gunungapi Aktif di Indonesia, edisi ke-2. Penerbit PT Citra Aji Parama,
Yogyakarta (2013b)
Kusumayudha, S.B., Pratiknyo, P., Riyanto, A.: Hidrokimia airtanah lereng selatan Merapi pasca
erupsi 2006. Jurnal Teknologi Mineral 22(2), 144–153 (2009)
Kusumayudha, S.B.: Merapi: Sebuah Matarantai Sejarah, SKH “Kedaulatan Rakyat”, 22
February 2006 (2006a)
Kusumayudha, S.B., Murwanto, H.: Ancient lake track records, and its influence to modern
hydrogeologic condition of the Borobudur National Park Area, Central Java, Indonesia. In:
Pros 34th Congress of International Association of Hydrogeologists, Beijing (2006b)
Mulyaningsih, S., Sampurno, S., Zaim, Y., Puradimaja, D.J., McGeehin, J., Bronto, S.: Very old
and young temple discoveries in Yogyakarta area: based on Volcano-tratigraphic study. In:
Proceeding Volcano International Gathering, Volcano: Life, Prosperity and Harmony,
Yogyakarta, 4–10 September 2006, pp. 102–114 (2006a)
Mulyaningsih, S., Sampurno, S., Zaim, Y., Puradimaja, D.J., Bronto, S.: Old building
discoveries, parts of lost civilization in Yogyakarta areas: in geologic point of view. In:
Proceeding Volcano International Gathering, Volcano: Life, Prosperity and Harmony,
Yogyakarta, 4–10 September 2006, pp. 125–133 (2006b)
Poesponegoro, M.D., Notosusanto, N.: Sejarah Nasional Indonesia II, Edisi Pemutakhiran,
Penerbit Balai Pustaka, xxii+536 p. (2008)
Van Bemmelen, R.W.: The Geology of Indonesia, Vol. IA, 732 p. Gov. Print. Office, The Hague
Martinus Nijhoff (1949)
Zen, M.T.: Enigma Merapi, Sarasehan Merapi dan Sejarah Mataram, UPN “Veteran”
Yogyakarta. www.digitalglobe.org, www.earthobservatory.sg, www.wikipedia.org/wiki/
Daftar_candi_di_Indonesia (2006)
Causal Analysis and Stability Evaluation
of Loess Landslide in Yili Region of Xinjiang -
A Case Study of Alar Village Landslide
Fei Ai1(&), Fan Zhou2, Wanlin Peng3, Jian Liu1, Xiuping Yan3,
and Pengfei Chen3
1
State Key Laboratory of Geomechanics and Geotechnical Engineering,
Institute of Rock and Soil Mechanics, Chinese Academy of Science,
Wuhan, China
aifei_2007@hotmail.com
2
Hubei Highway Intelligent Maintenance Technology Corporation Limited,
Hubei, China
3
Third Geological Brigade, Geological, Mineral Exploration and Development
Bureau of Xinjiang Uygur Autonomous Region, Korla, China
Abstract. Loess landslides are pervasive in the Yili region of Xinjiang pro-
vince. Loess landslides occur with some inherent conditions and external trig-
gers. Taking the Alar Village landslide in Yili as an example, first, detailed
investigation of the inherent conditions of landslides were investigated, such as
slope lithologies, geological and hydrogeological conditions. Then, physical and
mechanical properties of the landslide materials are measured in laboratory.
Experiment results showed that rainfall has a great influence on the physical and
mechanical parameters of loess. Under natural conditions, the internal friction
angle of loess is 24.5° and the cohesion is 28.4 kPa, while under saturated
condition, the mechanical parameters of loess decreased sharply, and the internal
friction angle is 16.5°, cohesion of 11.0 kPa. Combined with external triggering
factors such as rainfall and earthquake, landslide stability is evaluated by two-
dimensional limit equilibrium and finite element method. The simulation results
indicated that the factor of safety of slope is the lowest under the earthquake
condition. With the groundwater level gradually increased, the factor of safety of
slope gradually decreased. The research results can provide theoretical guidance
for landslide prevention and control in Yili loess area.
1 Introduction
The loess in the Yili River Valley of Xinjiang is widely distributed and the loess
landslide geohazard is the most typical type of geological hazards in the region. The
number of loess landslides accounts for 80% of the landslides in Xinjiang (Shi et al.
2018). The study of the formation mechanism and inducing factors of loess landslides
is an important theoretical basis to mitigate loess landslides (Ma 2015; Carey et al.
2017; Zhang et al. 2017). A lot of research work have been carried out in view of the
formation mechanism, inducing factors, deformation characteristics and mechanical
properties of loess on Yili loess landslides.

https://doi.org/10.1007/978-3-030-02032-3_9
Causal Analysis and Stability Evaluation of Loess Landslide 95
Zhang et al. (2004) analyzed the formation conditions of loess landslides in

Xinyuan Mountain area through geological, topographical and slope stress conditions.
Yin et al. (2009) studied the relationship of physical properties and dynamics of Yili
loess with landslides. An et al. (2010) analyzed the effects of precipitation and snow
melting, earthquakes, vegetation, and human engineering activities on the loess land-
slide in the Yili region. Wei et al. (2017) and Zhuang et al. (2018) studied the
mechanism of the landslide in Piliqing River of Yining and pointed out that the
infiltration of ice and snow was the direct cause of landslide. Wang and Mai (2016)
studied the formation mechanism and deformation characteristics of the Jialangpute
loess landslide while. Luan et al. (2016) performed a three-dimensional stability
analysis of the Jialangpute loess landslide. Wang et al. (2016) used the limit equilib-
rium and finite element strength reduction methods to calculate the landslide stability of
loess landslides under rainfall infiltration and earthquake conditions.
In this paper, detailed investigation of the inherent conditions of Alar Village loess
landslides were investigated, the physical and mechanical properties of the landslide
materials are measured in laboratory. Landslide stability is evaluated by two-
dimensional limit equilibrium and finite element method. The research results can
provide theoretical guidance for landslide prevention and control in Yili loess area.
2 Geological Conditions
2.1 Location, Climate and Hydraulic Information

The study area is located on the south bank of Mohur river of Alar Village, Gongliu
County, Yili region, Xinjiang with the geographical coordinates: longitude 82°42′
01.38′′–82°42′45.81′′ and north latitude 43°14′03.15′′–43°15′10.27′′, as shown in
Figs. 1 and 3.
Fig. 1. Location of the study area and the epicenters of earthquakes around the study area
96 F. Ai et al.
The study area belongs to the continental northern temperate climate. The tem-
perature difference between day and night is large. The summer heat is less and the
winter cold is also less, while the spring temperature rises fast and the autumn tem-
perature drops fast. The average annual temperature is around 8.3 °C, as shown in
Fig. 2.
Fig. 2. Precipitation, evaporation and temperature data of the study area
The precipitation in the study area increases in April and reaches the maximum in
July. The average annual precipitation is between 300 and 600 mm, and the average
annual evaporation is about 1450 mm, as shown in Fig. 2.
The Mohur river in the study area flows from the northern side of the study area to
the east-west direction, with an average annual runoff of 59.9 106 m3 and a total
catchment area of 213 km2, as shown in Fig. 3.
Fig. 3. Topographic map of the study area

2.2 Topography
The geomorphological unit of the study area is a hilly area, as shown in Figs. 3 and 4.
The topography of the area is undulating with high in the west and low in the west. The
elevation is between 1138–1522 m. The secondary landform types can be divided into
southern low mountains, loess platforms, landslide accumulation bodies and gently
inclined alluvial valley plains.
Fig. 4. Field photo of the study area
2.3 Lithology
The lithology of study area includes: the Quaternary Holocene landslide deposits
distributed on the old landslide deposits which is composed of silty soil, gravel sand,
and fully weathered mudstone mixed gravel layer, the Quaternary upper-Pleistocene
aeolian strata which is pale yellow to gray-yellow eolian loess showing a slightly
dense-medium-like, pore development, collapsibility, the mid-Pleistocene ice-water
sediments with lithology of pebbles and the Neogene brown red-brown mudstone, as
shown in Fig. 5.
Fig. 5. Profile map of the old landslide
2.4 Earthquake
The study area belongs to the western of the northern Tianshan earthquake zone. The
neo-tectonic movement is strong. According to the China earthquake parameters
zoning map, the basic seismic intensity in the study area is VIII degrees and the ground
98 F. Ai et al.
motion peak acceleration is 0.2 g. On November 1st, 2011, an earthquake measuring

6.0 on the Richter scale occurred at the junction of Gongliu County and Yining County
in Yili. The focal depth was 28 km and the epicenter was about 46 km from the study
area, as shown in Fig. 1.
3 Landslide Characteristics and Formation Mechanism

3.1 Spatial Morphology and Basic Characteristics
The morphology of the old landslide likes circle chair and slip surface is circular, with a
length of 700 m in the vertical direction, 200–600 m in the lateral width, an area of
about 35 104 m2, and a volume of more than 1000 104 m3. The original slope
angle of landslide is 30–35°, the length is 700 m, and the width is 200–600 m, as
shown in Figs. 3 and 4.
The back wall of the landslide is 60 m high, and the slope angle is 47–50°. The
upper part of the back wall is exposed. The trailing edge forms an arc-type tensile crack
with 245 m long, which roughly parallel to the back wall of the landslide. The tensile
crack formed a shifting surface with a displacement height of 30–50 cm.
3.2 Formation Mechanism

The landslide accumulation body is not formed by a one-time overall sliding but is a
traction-multiple sliding complex. The multiple slides mainly slipped along the contact
surface between the ice-water sediments and the fully weathered mudstone layer. The
slide soil is fully weathered mudstone, and the sliding surface is circular. The loess
landslide was judged to be an old landslide with multiple slides at the mudstone-loess
interface. The mechanism of old landslide is rotational slide and the unstable slope
above old landslide is prone to planar sliding.
4 Landslide Stability Analysis
4.1 Stability Calculation Model

The central axis of the old landslide is selected as the calculation profile, and the
direction of the profile is in the same direction as the main slide, as shown in Fig. 3.
The loads considered in the stability calculation of the landslide are mainly the body
weight of the landslide, groundwater, and seismic forces. Three calculation conditions
are considered: natural conditions with sliding body weight, natural + storm condition
with sliding body and osmotic pressure, natural + seismic condition with sliding body
weight and seismic force. The earthquake acceleration value is taken of 0.20 g. The
physical and mechanical tests of mudstone, gravel, silt and loess under natural and
saturated conditions were conducted respectively, and the physical and mechanical
parameters of rock and soil were obtained as shown in Table 1 and Fig. 6.
Table 1. Physical and geotechnical properties of the old landslide

Lithology State Weight Young’s Poisson’s Cohesion Friction
(kN/m3) modulus ratio (kPa) angle (°)
(MPa)
Loess Natural 20.5 75 0.39 28.4 24.5
Saturated 21.5 60 0.43 11.0 16.5
Silt Natural 16.9 70 0.39 30.4 22.6
Saturated 17.8 55 0.43 12.2 15.4
Gravel Natural 20.1 80 0.38 15 25
Saturated 21.1 64 0.42 13.5 22.5
Mudstone Natural 25.5 800 0.28 33.5 38.07
Saturated 25.7 600 0.31 10.0 28.5
Fig. 6. Direct shear test results of loess in natural and saturated conditions
4.2 Stability Calculation Results

Using two-dimensional limit equilibrium and two-dimensional finite element method to
calculate the factor of safety of the landslide in different working conditions, the results
are shown in Table 2. The calculation results show that under natural conditions, the
landslide factor of safety is slightly greater than 1.74 and is in stable state. When a
rainstorm or an earthquake is encountered, the factor of safety is less than 1.1 and is in
an unstable state. The slope may be instability and landslides in the snow-melting
season, infiltration of rainfall and earthquakes conditions.
Using the finite element method to calculate the horizontal displacement and ver-
tical displacement of the landslide under seismic condition, the results are shown in
Fig. 7. The calculation results show that under the earthquake conditions, the maxi-
mum horizontal displacement of the landslide is 0.27 m, which occurs in the back wall
of old landslide. The maximum vertical displacement of the landslide is 0.15 m, which
occurs in the surface part of the back wall of the old landslide.
100 F. Ai et al.
Table 2. Factor of safety results in different conditions by two methods

Conditions Factor of safety Stability
Two-dimensional limit Two-dimensional finite statuses
equilibrium method difference method
Natural 1.74 1.76 Stable
Rainfall 1.09 1.13 Basically
stable
Earthquake 1.03 1.06 Understable
Fig. 7. Contour plots of displacement data of old landslide simulated by two-dimension finite
element method: a. horizontal displacement, b. vertical displacement
4.3 Displacement Monitoring Results

Since 2010, a number of monitoring points have been laid at the slope of the trailing
edge of the old landslide to measure the horizontal and vertical displacement changes
of the slope, as shown in Fig. 3. As of the end of September 2013, the measured values
of horizontal displacement and vertical displacement of the P3 monitoring point are
shown in Fig. 8. The total horizontal displacement of the P3 point was 0.22 m and the
vertical displacement was 0.2 m. Overall, the horizontal displacement and the vertical
displacement of the P3 show a gradually increasing trend.
Fig. 8. Horizontal and vertical displacement data of the monitoring point P3: a. horizontal
displacement, b. vertical displacement
5 Discussion
5.1 The Influence of Topographic Conditions on Landslide Stability
The study area is located in low mountain with a slope angle of 25–40° and slope
orientation consistent with the rock formations. The continuous erosion of the water
flow of the Mohur river in the front of the old landslide provides a certain amount of
sliding force and air space conditions for the deformation of the landslide. It is not
conducive to the stability of landslides and is likely to trigger landslides.
The back wall of the old landslide is 60 m high with a gradient of 47–50°. The
upper part of the back wall is exposed and the traces of rain erosion are locally visible.
The trailing edges of back wall gradually forms arc-type tensile cracks. Landslides may
occur due to instability when the snow melting season, rainfall infiltration and earth-
quake occur.
5.2 The Influence of Earthquakes on Landslide Stability

Earthquakes are one of important triggering factors of geological hazards. There are
two aspects to the inducement of landslides: one is the triggering effect. The earthquake
causes the slippage of the slope near the critical stable state. The second is the
102 F. Ai et al.
cumulative effect. The earthquake action makes the slope structure loose and slopes are
prone to cracks which create the conditions for precipitation infiltration and promoting
unstable failure of slopes. The peak acceleration of ground motions in the study area is
0.20 g, and the basic intensity is VIII degrees. According to statistics, the number of
landslides in Area VIII degrees of Gongliu County accounts for 84% of the total
number of landslides of Gongliu County. It is indicated that the seismic activity is
closely related to the stability and spatial distribution of loess landslide. Affected by the
6.0 earthquake occurred on November 1st, 2011, the horizontal displacement and
vertical displacement of the monitoring point P3 showed a sharp increase, which was
increased by 0.1 m and 0.05 m, respectively, as shown in Fig. 8. It shows that the
earthquake caused by the cumulative deformation of landslides.
5.3 The Influence of Groundwater on Landslide Stability

The infiltration of precipitation and snow melting water are not only increasing the
gravity and infiltration pressure of the landslide, but also significantly reduce the shear
strength of the soil and rock of landslide. When the sliding force is greater than the anti-
sliding force, landslides are easily induced. From the physical and mechanical test
results of rock and soil, it can be found that the internal friction angle and cohesion of
rock and soil are significantly lower under saturated conditions than under natural
conditions. Under natural conditions, the internal friction angle of loess is 24.5° and the
cohesion is 28.4 kPa, while under saturated condition, the mechanical parameters of
loess decreased sharply, and the internal friction angle is 16.5°, cohesion of 11.0 kPa,
as shown in Fig. 6. Compared with natural conditions, the internal friction angle and
cohesion of soil under saturated conditions are reduced by 33% and 62%, respectively.
The factor of safety of slope under different groundwater conditions is shown in Fig. 9.
The calculation results show that the factor of safety of the slope gradually decreases as
the groundwater table rises.
Fig. 9. The relationship between the groundwater table and factor of safety of slope
6 Conclusions
Based on the geological conditions, laboratory test and slope stability analysis, the
causal analysis and stability evaluation of loess landslide in Yili region of Xinjiang are
analyzed. The conclusions are as follows:
The detailed investigation of the inherent conditions of landslides were investi-
gated, such as slope lithologies, geological results and hydrogeological conditions. It
found that the back wall of the old landslide is 60 m high with a gradient of 47–50°.
The trailing edges of back wall gradually forms arc-type tensile cracks. Landslides may
occur due to instability when the snow melting season, rainfall infiltration and earth-
quake occur.
The experiment results showed that rainfall has a great influence on the physical
and mechanical parameters of loess. Under natural conditions, the internal friction
angle of loess is 24.5° and the cohesion is 28.4 kPa, while under saturated condition,
the mechanical parameters of loess decreased sharply, and the internal friction angle is
16.5°, cohesion of 11.0 kPa.
Landslide stability is evaluated by two-dimensional limit equilibrium and finite
element method. The simulation results indicated that the factor of safety of slope is the
lowest under the earthquake condition. With the groundwater level gradually increased,
the factor of safety of slope gradually decreased.
Affected by the 6.0 earthquake occurred on November 1st, 2011, the horizontal
displacement and vertical displacement of the monitoring point P3 showed a sharp
increase, which was increased by 0.1 m and 0.05 m, respectively. The finite element
numerical simulation of the slope displacement values is basically consistent with the
monitoring displacement values. The research results can provide theoretical guidance
for landslide prevention and control in Yili loess area.
Acknowledgments. The financial supports from the National Natural Science Foundation of
China (No. 41602320) are gratefully acknowledged. We also like to acknowledge the reviewers
for their constructive remarks.
References
An, H., Liu, P.: Genesis and influencing factors of loess landslides in Yili region in Xinjiang.
Geol. Hazards Environ. Prot. 21(3), 22–25 (2010)
Carey, J.M., et al.: Dynamic liquefaction of shear zones in intact loess during simulated
earthquake loading. Landslides 14(3), 789–804 (2017)
Luan, Z., et al.: Analysis on geological features and stability of Jialangpute loess landslide
group. J. Shihezi Univ.: Nat. Sci. 34(5), 643–648 (2016)
Ma, L.: The effect of moisture content on the loess physical properties. Adv. Mater. Res. 1065,
81–84 (2015)
Shi, J.S., et al.: Neogene clay and its relation to landslides in the southwestern Loess Plateau.
China. Environ. Earth Sci. 77(5), 204 (2018)
Wang, X., Mai, Z.: Sliding mechanism and deformation characteristics of typical large loess
landslides in Yili Xinjiang. J. Water Resour. Arch. Eng. 14(4), 195–200 (2016)
104 F. Ai et al.
Wang, Y., et al.: Stability analysis and case study of rainfall and disturbance on loess slope. In:
International Conference on Smart City and Systems Engineering, pp. 64–67 (2016)
Wei, Y., et al.: Analysis of formation mechanism of Piliqinghe landslide in Yining County,
Xinjiang Province. Chin. J. Geol. Hazard Control. 28(4), 23–26 (2017)
Yin, G., et al.: Physical index, dynamic property and landslide of Yili loess. Arid Geogr. 32(6),
899–905 (2009)
Zhang, H., et al.: Forming conditions and prevention treatment of loess landslide in the Xinyuan
mountain area, Xinjiang. Xinjiang Geol. 22(3), 233–237 (2004)
Zhang, Z., et al.: Seismic performance of loess-mudstone slope in Tianshui-Centrifuge model
tests and numerical analysis. Eng. Geol. 222, 225–235 (2017)
Zhuang, M., et al.: Type and characteristics of loess landslides in Piliqing River, in Yili of
Xinjiang Uygur autonomous region. Chin. J. Geol. Hazard Control. 29(1), 54–59 (2018)
Artificial Neural Networks for Rock and Soil
Cutting Slopes Stability Condition Prediction
Joaquim Tinoco1,2(&), António Gomes Correia1, Paulo Cortez2,

and David Toll3
1
ISISE – Institute for Sustainability and Innovation in Structural Engineering,
School of Engineering, University of Minho, Guimarães, Portugal
jtinoco@civil.uminho.pt
2
ALGORITMI Research Center/Department of Information Systems,
School of Engineering, University of Minho, Guimarães, Portugal
3
School of Engineering and Computing Sciences, University of Durham,
Durham, UK
Abstract. This study aims to develop a tool able to help decision makers to find
the best strategy for slopes management tasks. It is known that one of the main
challenges nowadays for every developed or countries undergoing development
is to keep operational under all conditions their transportations infrastructure.
However, due to the network extension and increased budget constraints such
challenge is even more difficult to accomplish. Keeping in mind the strong
impact of a slope failure in the transportation infrastructure it is important to
develop tools able to help minimizing this situation. Accordingly, and in order
to achieve this goal, the high flexible learning capabilities of Artificial Neural
Networks (ANNs) were applied in the development of a classification tool
aiming to identify the stability condition of a rock and soil cutting slopes,
keeping in mind the use of information usually collected during routine
inspections activities (visual information) to feed them. For that, it was followed
a nominal classification strategy and, in order to overcome the problem of
imbalanced data, three training sampling approaches were explored: no
resampling, SMOTE (Synthetic Minority Over-sampling Technique) and
Oversampling. The achieved results are presented and discussed, comparing the
achieved performance for both slope types (rock and soil cuttings) as well as the
effect of the sampling approaches. An input-sensitivity analysis was applied,
allowing to measure the relative influence of each model attribute.
1 Introduction
A nowadays challenge concerning to the transportation network, mainly for every

developed or countries undergoing development that have invested and keep investing
to build a safe and functional transportation network, is how to keep it operational
under all conditions, keeping in mind its extension and the increased budget limitation
for maintenance and repair tasks. Thus, taken into account the key importance of the
transportation system in modern societies, it is fundamental to develop new tools able
to help in its management.

https://doi.org/10.1007/978-3-030-02032-3_10
106 J. Tinoco et al.
Keeping in mind the strong impact of a slope failure in the transportation infras-
tructure, namely for the railway it is important to develop tools able to help minimizing
this situation.
Over time, several efforts have been made toward the development of a system to
detect slope failures. However, most of the systems were developed for natural slopes,
presenting some limitations when applied to engineered (human made) slopes. In
addition, they have limited applicability as most of them were developed based on
particular case studies or using small databases. Furthermore, another aspect that can
limit its applicability is related with the information required to feed them, such as data
taken from complex tests or from expensive monitoring systems. Pourkhosravani and
Kalantari (2011) summarized in their work some of the current methods for slope
failure detection, which were grouped into Limit Equilibrium (LE) methods, Numerical
Analysis methods, Artificial Neural Networks and Limit Analysis methods. There are
also approaches based on finite elements methods (Suchomel et al. 2010), reliability
analysis (Sivakumar and Murthy 2005; Husein Malkawi et al. 2000), as well as some
methods making use of data mining (DM) algorithms (Cheng and Hoang 2016;
Ahangar-Asr et al. 2010; Yao et al. 2008; Kang et al. 2017; Kang and Li 2016; Suman
et al. 2016). More recently, a new flexible statistical system was proposed by Pinheiro
et al. (2015), based on the assessment of different factors that affect the behavior of a
given slope, which are weighted in order to calculate a final indicator of the slope
stability condition.
In summary, most of the approaches so far proposed share the main limitations,
which are related with its applicability domain or dependency on information that is
difficult to obtain. In fact, the assessment of the stability condition of given slope is a
multi-variable problem characterized by a high dimensionality.
Artificial Neural Networks (ANNs) are one of the most well known Data Mining
(DM) algorithms, which have been applied with success in different knowledge
domains, such as web search, spam filters, recommender systems, and fraud detection
(Domingos 2012). Also in civil engineering field, several applications can be found.
For example, ANNs were applied in the study of physical and mechanical properties of
jet grouting columns (Tinoco et al. 2014, 2016). Indeed, the high learning capabilities
of this algorithm give it the ability to model complex nonlinear mappings. Thus, in this
work the high flexible learning capabilities of Artificial Neural Networks (ANNs) were
used to predict the stability condition of rock and soil cutting slopes according to a pre-
defined classification scale based on four levels (classes). Moreover, one of the
underlying premises of this work is to use only information that can be easily obtained.
For that, more than fifty variables related with data collected during routine inspections
as well as geometric, geological and geographic data were used to feed the models.
With this methodology it is intended to identify critical zones for which more detailed
information can then be obtained in order to perform more detailed stability analysis.
Such novel approach is intended to support railway network management companies to
allocate the available funds to the priority assets according to their stability condition.
Artificial Neural Networks for Rock and Soil Cutting Slopes 107
2 Methodology
Data characterization. To fit the proposed models for stability condition identifica-

tion, from this point referred to as EHC (Earthwork Hazard Category (Power et al.
2016)), of rock and soil cutting slopes two database were compiled respectively. Both
databases, containing information collected during routine inspections and comple-
mented with geometric, geological and geographic data of each slope, were gathered by
Network Rail workers and are concerned with the railway network of the UK. For each
slope a class of the EHC system was defined by the Network Rail Engineers based on
their experience/algorithms (Power et al. 2016), which will be assumed as a proxy for
the real stability condition of the slope for year 2015. The EHC system comprises 4
classes1 (“A”, “B”, “C” and “D”) where “A” represents a good stability condition and
“D” a bad stability condition. In other words, the expected probability of failure is
higher for class “D” and lower for class “A”.
Both databases contain a significant number of records. The rock slopes database
comprises 5945 records, while the soil cutting slopes database is bigger, having 10928
records available. Figure 1 plots the distribution of EHC classes for each database.
From their analysis, it is possible to observe a high asymmetric distribution (imbal-
anced data), particularly for the rock cutting slopes database. In fact, more than 86% of
the rock slopes are classified as “A”. Although this type of asymmetric distribution,
where most of the slopes present a low probability of failure (class “A”), is normal and
desirable from the safety point of view and slope network management, it can represent
an important challenge for data-driven models learning, as detailed in next section. The
proposed models for EHC identification of rock and soil cutting slopes were fed with
more than fifty variables normally collected during routine inspections and comple-
mented with geometric, geographic and geological information. To be precise, 65
variables were used in the rock slopes study and 51 variables in soil cutting slopes.
Since the number of analyzed variables is high (65/51), just a few examples of the
variables used to feed the models are here enumerated: height, slope angle, presence of
rock outcrops, animal activity, presence of boulders, ground cover, rock type, dan-
gerous trees, number of root balls, rock strength, etc.
Modeling. In this work, ANNs were trained to predict EHC of rock and soil cutting
slopes. ANNs, although not new, they are supported in a strong background and
experience. Indeed, they have been applied in the past with high success in different
knowledge domains including in civil engineering (Chou et al. 2016; Correia et al.
2013). There are also some examples of ANNs applications in slope stability analysis
(Wang et al. 2005; Cheng et al. 2012).
ANNs are learning machines that were initially inspired in functioning of the
human brain (Keniget et al. 2001). The information is processed using iteration among
several neurons. This technique is capable of modeling complex non-linear mappings
and is robust in exploration of data with noise. In this study we adopt the multilayer
1
The original EHC system comprised 5 levels (A, B, C, D and E) (Power et al. 2016). However, and due
to the reduced number of slopes classified as E, classes D and E were combined in one, named as D.
EHC: A B C D
6297
6000
5124
Number of records
4000
3076
2000
1426
569
215
129
37
0
Rock Soil
Slope type
Fig. 1. Rock and soil cutting slopes data distribution by EHC classes
perceptron that contains only feedforward connections, with one hidden layer con-
taining H processing units. Because the ANN performance is sensitive to H (a trade-off
between fitting accuracy and generalization capability), a grid search of {0; 2; 4; 6; 8}
was adopted under an internal (i.e. applied over training data) three fold cross vali-
dation during the learning phase to find the best H value. Under this grid search, the
H value that produced the lowest MAE (Mean Absolute Error) was selected, and then
the ANN was retrained with all of the training data. The neural function of the hidden
nodes was set to the popular logistic function 1/(1 + e−x).
The problem of EHC determination was approached following a nominal classi-
fication strategy, where the network calculates a probability for each EHC level and
then is selected the class corresponding to the highest probability.
In addition, and in order to minimize the effect of the imbalanced data (see Fig. 1),
Oversampling (Ling and Li 1998) and SMOTE (Synthetic Minority Oversampling
Technique) (Chawla et al. 2002) approaches were applied over the training data before
fitting the models. When approaching imbalanced classification tasks, where there is at
least one target class label with a smaller number of training samples when compared
with other target class labels, the simple use of a soft computing training algorithm will
lead to data-driven models with better prediction accuracies for the majority classes and
worst classification accuracies for the minority classes. Thus, techniques that adjust the
training data in order to balance the output class labels, such as Oversampling and
SMOTE, are commonly used with imbalanced datasets. In particular, Oversampling is
a simple technique that randomly adds samples (with repetition) of the minority classes
to the training data, such that the final training set is balanced. SMOTE is a more
sophisticated technique that creates “new data” by looking at nearest neighbors to
establish a neighborhood and then sampling from within that neighborhood. It operates
on the assumptions that the original data is similar because of proximity. We note that
the different sampling approaches were applied only to training data, used to fit the
data-driven models, and the test data (as provided by the 5-fold procedure) was kept
without any change.
For models evaluation and comparison, three classification metrics were adopted:
recall, precision and F1-score (Hastie et al. 2009). The recall measures the ratio of how
many cases of a certain class were properly captured by the model. In other words, the
recall of a certain class is given by TruePositives=ðTruePositives þ FalseNegativesÞ.
On the other hand, the precision measures the correctness of the model when it predicts
a certain class. More specifically, the precision of a certain class is given by
TruePositives=ðTruePositives þ FalsePositivesÞ. The F1-score was also calculated,
which represent a trade-off between the recall and precision of a class. The F1-score
correspond to the harmonic mean of precision and recall, according to the following
expression:
precision recall
F1score ¼ 2 ð1Þ
precision þ recall
For all three metrics, the higher the value, the better are the predictions, ranging
from 0% to 100%.
The generalization capacity of the models was accessed through a 5-fold cross-
validation approach under 20 runs (Hastie et al. 2009). This means that each modeling
setup is trained 5 10 ¼ 100 times. Also, the three prediction metrics are always
computed on test unseen data (as provided by the 5-fold validation procedure).
All experiments were conducted under the R statistical environment (R Team
2009). ANN algorithm was trained using the rminer package (Cortez 2010), which
facilitates its implementation, as well as different validation approaches such as the
cross validation adopted in this work.
Following are presented and discussed the achieved performance in EHC determination
of both rock and soil cutting slopes based on ANN algorithm, as well as the effect of
the three training sampling approaches explored: Normal (no resampling), OVERed
(Oversampling) and SMOTEd (SMOTE).
Figure 2 gives an overview of ANNs models performance in rock and soil cutting
slopes, based on recall, precision and F1-score.
Concerning to rock cutting slopes study (left side of Fig. 2) the achieved perfor-
mance is somewhat lower independently of the resampling approach applied. Although
a very high performance is observed for class “A” (F1-score higher than 95%), for class
“C” and particularly for class “D”, all models evidence clear difficulties in predicting
such classes correctly. In fact, and using F1-score as reference, the best performance in
identification of slopes of class “D” is lower than 14% (see Fig. 2) which was achieved
by applying SMOTE resampling approach. From Fig. 3a analysis, which plots the
Approach: Normal OVERed SMOTEd
Rock Soil
100 95.44 93.95 93.01
88.77 86.35 85.94
75
62.13
F1-score
59.94
56.82 55.91
50.93 51.35 48.95 51.53
48.62
50
36.27 35.91
30.49 31.33
26.84
25 21.67
12.77 13.37
5.77
0
98.01 98.5
100 94.66
91.13 91.68
87.23
Metric value
75
Precision
60.36 59.21
55.02 54.49
49.06 49.77 51.48
50 41.27 42.57
38.36 39.22
33.47 33.71
26.14
25 21.63
13.71 12.7 10.89
0
100 96.23
90.21 90.36
88.1
82.05 80.87
75 67.96 67.6 66.59
64.01 63.77
58.75 56.78
Recall
52.95
50 45.61 46.07
39.58 36.58 38.41
25 20.4
17.3
12.84 14.53
3.65
0
A B C D A B C D
EHC
Fig. 2. ANN performance comparison in EHC determination of both rock and soil cutting
slopes based on recall, precision and F1-score
relation between observed and predicted EHC values of rock cutting slopes by applying
oversampling approach (best overall fit), it is clear the model difficulties in correctly
predicting class “C” and particularly class “D”, for which the expected probability of
failure is higher. As shown, only around 12% of rock cutting slopes classified as “D”
are correctly identified, which represents a low performance, namely when compared
with soil cutting slopes study as following discussed. Overall, these results show that
the methodology applied for EHC determination of rock cutting slopes needs future
development in order to overcome this gap.
Relating to soil cutting slopes study, a very promising performance is observed, as
shown in Fig. 2 (right side). For example, soil cutting slopes of class “A” can be
correctly identified, either with or without resampling. Also for classes “B” and “C” a
promising performance is also observed, with an F1-score around 55%. Concerning the
class “D”, although an F1-score lower than 36% was achieved, the obtained value for
recall metric around 57% shows a good performance for class “D” identification.
Analyzing Fig. 3b that shows the relation between observed and predicted EHC values
according to the best overall fit (SMOTEd), one can see that the model performance is
very promising. Indeed, when a SMOTE resampling approach is applied, ANN algo-
rithm is able to predict correctly around 57% of soil cutting slopes of class “D”, which
represents a very promising performance if we take into account that this is the
minority class. For class “C”, although the accuracy is not so high (around 40%), when
not predicted as “C” they are classified as belonging to the closest class, that is, “B” or
ANN :: OVERed - Nominal Classification ANN :: SMOTEd - Nominal Classification
100000
75000
75000
EHC predicted EHC predicted

Frequency
Frequency
50000 A A
B 50000 B
C C
D D
25000
25000
0 0
A B C D A B C D
a) EHC Experimental
b) EHC Experimental
Fig. 3. ANN performance: (a) Rock cutting slopes following an OVERed approach; (b) Soil
cutting slopes following a SMOTEd approach
“D”. This type of misclassification is also observed for classes “A”, “B” and “D”,
which can be interpreted as positive point. Concerning to classes “A” and “B”, the
ANN model was able to identify it very accurately.
Comparing the achieved results of rock and soil cutting slopes studies, the proposed
models for soil cutting slopes are more effective, namely in the identification of classes
“C” and “D” for which the probability of failure is higher (see Fig. 2). A possible
explanation for the lower performance, namely for classes “C” and “D” of rock cutting
slopes could be related with the EHC classes being assumed as representative of the
real stability condition of each slope. Indeed, analyzing the number of slope failures by
EHC class for rock slopes there are some indications that the classification attributed to
each rock slope could lack of some accuracy as reported in the work of Power et al.
(2016), which used the same source of information. It would be expected that most of
the failures would occur in slopes of classes “C” and mainly “D”. However, for rock
slopes such behavior is not observed as reported on Power et al. (2016). In fact, the
number of failures for each EHC class is almost constant from classes “A” to “D”,
particularly when compared with soil cuttings. For example, the number of failures
observed in rock cutting slopes of class “C” is only twice higher when compared to
class “A”. This observation shows that the defined classes for rock slopes have a poor
correlation with actual failures.
These results show that a deeper data analysis is required, particularly in the study
of rock cutting slopes. For example, the number of variables taken as model attributes
might be too high and may be influencing the generalization performance of the
models.
Thus, and in order to better understanding the proposed models, the relative
importance of each model input was measured based on a sensitive analysis. For that,
the methodology proposed by Cortez and Embrechts (2013) was applied. Accordingly,
Fig. 4 shows the relative importance of the 20 most relevant variables based on the two
models that achieved the overall best performance in EHC determination of rock (left
side) and soil (right side) cutting slopes (see Fig. 3). From their analysis, it is observed
that the height of the slope is one of the most relevant variables affecting the slope
stability condition, either for rock or soil cutting slopes. On the other hand, and par-
ticularly for rock slopes, it is observed that the first seven most relevant variables are
responsible for around 70% of the total influence.
ANN :: OVERed - Nominal Classification [Importance] ANN :: SMOTEd - Nominal Classification [Importance]
RS Actual Height 17.5 End Height 5.5
US Actual Height 15.4 Actual Height1 5.3
US Actual Angle 10.1 Start Height 5.1
RS Dangerous Trees Number 8.7 Actual Hyp3 4.3
RS Actual Hyp 7.2 Cutting Cess Drainage 4.2
LS Actual Height 5.6 Max Height 3.9
US Actual Hyp 5.4 Validate Cracking 3.7
RS Azimuth 4.2 Actual Height2 3.7
Cess Safe 3.9 Composition Crest 3.6

Input Variable
Input Variable
RS Root Balls Number 3.4 Adjacent Catch Area 3.5
LS Actual Angle 3.2 Actual Height3 3.3
LS Actual Hyp 1.9 Boulders Present 2.9
Pot Failure On Slope 1.7 Min Height 2.6
RS Height 1.5 Tree Cover 2.6
Upper Slope 1.2 Validate Instability 2.5
RS Length 1 Actual Crest Width 2.4
End Northing 0.8 Actual Hyp2 2.4
Drainage Problems 0.6 Actual Slope To Track 2.3
Cess Stand Off 0.6 Actual Hyp1 2.3
LS Angle 0.6 Catchment Surface 2.1
0 5 10 15 0 2 4
a) Relative Importance (%)

b) Relative Importance (%)
Fig. 4. Relative importance bar plot of the 20 most relevant variables according to ANN models
in EHC determination: (a) Rock cutting slopes following an OVERed approach; (b) Soil cutting
slopes following a SMOTEd approach
These results, together with achieved performance, suggest a need of revision of the
variables used as model attributes. Accordingly, and as a future works, it is intended to
apply a more sophisticated feature selection method in order to improve models per-
formance. In particular, it will be explored a multi-objective evolutionary computation
method that simultaneously maximizes prediction performance and minimizes the
number of inputs used.
As a final observation, and considering the overall performance of all models, it
should be underlined the potential of ANNs, or even other soft computing algorithms,
in EHC determination of rock and soil cutting slopes.
Acknowledgments. This work was supported by FCT - “Fundação para a Ciência e

a Tecnologia”, within ISISE, project UID/ECI/04029/2013 as well Project Scope:
UID/CEC/00319/2013 and through the post-doctoral Grant fellowship with reference
SFRH/BPD/94792/2013. This work was also partly financed by FEDER funds through the
Competitivity Factors Operational Programme - COMPETE and by national funds through FCT
within the scope of the project POCI-01-0145-FEDER-007633. This work has been also
supported by COMPETE: POCI-01-0145-FEDER-007043. A special thanks goes to Network

Rail that kindly made available the data (basic earthworks examination data and the Earthworks
Hazard Condition scores) used in this work.
References
Ahangar-Asr, A., Faramarzi, A., Javadi, A.: A new approach for prediction of the stability of soil
and rock slopes. Eng. Comput. 27(7), 878–893 (2010)
Chawla, N., Bowyer, K., Hall, L., Kegelmeyer, W.: Smote: synthetic minority oversampling
technique. J. Artif. Intell. Res. 32, 321–357 (2002)
Cheng, M., Hoang, N.: Slope collapse prediction using bayesian framework with k-nearest
neighbor density estimation: case study in Taiwan. J. Comput. Civ. Eng. 30(1), 04014116
(2016)
Cheng, M., Roy, A., Chen, K.: Evolutionary risk preference inference model using fuzzy support
vector machine for road slope collapse prediction. Expert Syst. Appl. 39(2), 1737–1746
(2012)
Chou, J., Yang, K., Lin, J.: Peak shear strength of discrete fiber-reinforced soils computed by
machine learning and metaensemble methods. J. Comput. Civ. Eng. 04016036 (2016)
Cortez, P.: Data mining with neural networks and support vector machines using the R/rminer
tool. In: Perner, P. (ed.) Advances in Data Mining: Applications and Theoretical Aspects.
ICDM 2010. LNAI, vol. 6171, pp. 572–583. Springer (2010)
Cortez, P., Embrechts, M.: Using sensitivity analysis and visualization techniques to open black
box data mining models. Inf. Sci. 225, 1–17 (2013)
Domingos, P.: A few useful things to know about machine learning. Commun. ACM 55(10), 78–
87 (2012)
Correia, A.G., Cortez, P., Tinoco, J., Marques, R.: Artificial intelligence applications in
transportation geotechnics. Geotech. Geol. Eng. 31(3), 861–879 (2013)
Hastie, T., Tibshirani, R., Friedman, J.: The Elements of Statistical Learning: Data Mining,
Inference, and Prediction, 2nd edn. Springer-Verlag, New York (2009)
Husein Malkawi, A., Hassan, W., Abdulla, F.: Uncertainty and reliability analysis applied to
slope stability. Struct. Saf. 22(2), 161–187 (2000)
Kang, F., Li, J.: Artificial bee colony algorithm optimized support vector regression for system
reliability analysis of slopes. J. Comput. Civ. Eng. 04015040 (2016)
Kang, F., Li, J.S., Wang, Y., Li, J.: Extreme learning machine based surrogate model for
analyzing system reliability of soil slopes. Eur. J. Environ. Civ. Eng. 21(11), 1341–1362
(2017)
Kenig, S., Ben-David, A., Omer, M., Sadeh, A.: Control of properties in injection molding by
neural networks. Eng. Appl. Artif. Intell. 14(6), 819–823 (2001)
Ling, C., Li, C.: Data mining for direct marketing: problems and solutions. In: KDD 98, pp. 73–
79 (1998)
Pinheiro, M., Sanches, S., Miranda, T., Neves, A., Tinoco, J., Ferreira, A., Correia, A.G.: A new
empirical system for rock slope stability analysis in exploitation stage. Int. J. Rock Mech.
Min. Sci. 76, 182–191 (2015)
Pourkhosravani, A., Kalantari, B.: A review of current methods for slope stability evaluation.
Electron. J. Geotech. Eng. 16 (2011)
Power, C., Mian, J., Spink, T., Abbott, S., Edwards, M.: Development of an evidence-based
geotechnical asset management policy for network rail, Great Britain. Procedia Eng. 143,
726–733 (2016)
Sivakumar Babu, G., Murthy, D.: Reliability analysis of unsaturated soil slopes. J. Geotech.
Geoenviron. Eng. 131, 1423–1428 (2005)
Suchomel, R., et al.: Comparison of different probabilistic methods for predicting stability of a
slope in spatially variable c–u soil. Comput. Geotech. 37(1), 132–140 (2010)
Suman, S., Khan, S., Das, S., Chand, S.: Slope stability analysis using artificial intelligence
techniques. Nat. Hazards 84(2), 727–748 (2016)
Team, R.: R: A language and environment for statistical computing. R Foundation for Statistical
Computing, Vienna, Austria (2009). http://www.r-project.org/
Tinoco, J., Correia, A.G., Cortez, P.: Support vector machines applied to uniaxial compressive
strength prediction of jet grouting columns. Comput. Geotech. 55, 132–140 (2014)
Tinoco, J., Correia, A.G., Cortez, P.: Jet grouting column diameter prediction based on a
datadriven approach. Eur. J. Environ. Civ. Eng. 1–22 (2016)
Wang, H., Xu, W., Xu, R.: Slope stability evaluation using back propagation neural networks.
Eng. Geol. 80(3), 302–315 (2005)
Yao, X., Tham, L., Dai, F.: Landslide susceptibility mapping based on support vector machine: a
case study on natural slopes of Hong Kong, China. Geomorphology 101(4), 572–582 (2008)
On the Development of Ground-Based
and Drone-Borne Radar System
Tomonori Deguchi(&), Tomoyuki Sugiyama,

and Munemaru Kishimoto
Nittetsu Mining Consultants Co., Ltd., Tokyo, Japan

deguchi@nmconsults.co.jp
Abstract. The interferometric SAR (InSAR) technique is an established ana-

lytical tool designed to analyze phase information provided by satellite-mounted
synthetic aperture radar (SAR) to provide precise measurements of target dis-
placement to a centimeter level of precision. The InSAR technique has already
reached the practical stage in the measurement of ground subsidence in urban
areas and in crustal deformation analysis associated with earthquakes and vol-
canic activities. The authors are currently developing radar systems for on-
ground installation and mounting on drones. InSAR analysis can also be applied
to the data acquired by these systems. We ultimately aim to develop tech-
nologies that will allow us to monitor the stability of opencast mine slopes,
landslide slopes, and aging infrastructures.
Measurements taken by ground-based radar systems can be conducted at all
times automatically, without human operators. Measurement data can be accu-
mulated at intervals of several minutes to several tens of minutes. Millimeter
wave radar is internationally recognized as a collision prevention technology for
automobiles. Our group has used this technology to develop a millimeter wave
radar system capable of transmitting radio waves at a bandwidth (76 to 77 GHz)
approved by the regulatory authorities. In field tests, this radar was verified to
detect targets remotely from a distance of about 1.6 km and to measure dis-
placement with an accuracy of about 0.2 mm. In a long-term demonstration test
conducted inside an open-pit mine, the radar-collected data clearly captured the
precursor phenomena of the slope failure. The radar also successfully detected
anomalies corresponding to consolidation subsidence of sediment deposited by
slope failure.
The drone-borne radar we are developing will be used for periodical obser-
vations to be carried out several times a year. We designed the system to
transmit the microwave band (12.8 to 12.95 GHz) over a longer wavelength
than the millimeter wave band, as the phase information used for InSAR
analysis becomes unstable with time. The frequencies of the transmitted radio
waves adopted in this study, including those transmitted by the ground-based
radar, were selected to comply with the Radio Law of Japan. The ground-based
radar has reached a stage where it can be deployed practically in the field. Real-
life use of the drone-borne radar has yet to be realized, though a prototype
version of the system is now complete. We plan to develop algorithms for
synthetic aperture processing and interferometric analysis in the future.

https://doi.org/10.1007/978-3-030-02032-3_11
116 T. Deguchi et al.
1 Introduction
Our group is developing two radar systems, a ground-based type and drone-borne type.
By applying the InSAR technique to the data measured with these systems, we aim to
develop technologies for the monitoring of opencast mine slopes, landslide slopes, and
aging infrastructure. In this paper we report the current status of device development
and the results of field examinations to evaluate the performance of the systems.
2 Development of Ground-Based Radar System
Our ground-based radar system performs distance measurement by the conventional

FM-CW method. The specification is shown in Table 1. The radar system for this
device employs a real aperture and rotates in both the horizontal and vertical directions
(Fig. 1). The system can acquire two-dimensional images while repeatedly transmitting
and receiving millimeter waves when rotated on a rotation table installed on a base such
as a tripod. Image resolution is weakened in long-distance measurements by this real
aperture radar, as the 3 dB beam width of the antenna coincides with the azimuth
resolution. There is little concern, however, that the temporary vibrations issued by the
rotations of the stepping motor will hinder measurement, as no synthetic aperture
processing is required. This latter feature makes it possible to easily establish a highly
sustainable measurement system. The gain of the Cassegrain antenna is determined by
the antenna diameter and wavelength. With a millimeter wave system, a high gain can
be easily secured with only a small antenna diameter. This feature enables long-
distance measurement with lower power consumption.
Table 1. Specification of the ground-based radar system using millimeter radio wave.
Frequency 76.5 GHz
Wavelength 3.92 mm
Band width 1 GHz (adjustable)
Modulation FM-CW
Radar system Real aperture
Antenna type Cassegrain (20 cm)
Antenna power 10 mW (10 dBm)
Antenna gain 40 dBi
Range resolution 0.30 m
Azimuth resolution 1.23 degree
A/D converter 16 bits
Weight 10 kg
Accuracy of displacement less than 0.2 mm
The images in Fig. 2 show the details of a demonstration test performed on a slope
formed in an opencast mine. Figure 2(a) shows the landscape observed from the site of
On the Development of Ground-Based and Drone-Borne Radar System 117
Fig. 1. Conceptual diagram of the ground-based radar system developed.
radar installation, with radio waves transmitted to cover the area framed by the broken
red line. Figures 2(b) and (c) are a high-resolution photograph acquired from Google
Earth and a power radar image obtained by cutting out the same range, respectively.
This demonstration test clearly confirmed that the system could detect a rock slope
1.58 km away from the radar.
Fig. 2. An example of a two-dimensional radar image acquired of the opencast mine.

Although not displayed on the two-dimensional image shown in Fig. 2(c), the
phase information is recorded in the actual raw data, which enables the measurement of
minute displacement by applying the InSAR technique. We installed our millimeter
wave radar device inside the open pit mine and conducted a long-term demonstration
test over the period of 7 months. A small-scale slope failure occurred accidentally
within the observation range in the course of the test. The system recorded the ground
levels before and after the slope failure, reporting the displacement in between as
reported hereunder.
The photographs in Fig. 3 show the slope before and after the slope failure
occurred. The three-dimensional map in Fig. 4(a) displays the displacement measure-
ment results recorded over the course of about one month before the slope failure. The
green shading in the figure indicates no change; warm colors indicate displacement
toward the radar; and cold colors indicate displacement away from the radar. Warm
color and cold color anomalies are clearly detected in pairs in the area framed by the
solid black line. No slope failure has yet occurred at this stage, so we can infer that the
precursor phenomena was captured. The maximum displacement rate measured before
slope failure was approximately 4 mm/day.
Fig. 3. Pictures before and after slope failure.
From this measurement result we infer that as the surface layer gradually moved
toward the bottom of the slope, the volume at the lower part increased, which was
detected by the system anomalously as displacement toward the radar. Likewise, the
system anomalously detected the decrease in the volume of the upper part as dis-
placement away from the radar. The cumulative displacement of the precursor phe-
nomena exceeded 30 mm, though it might have been undetectable by the naked eye
without careful observation.
A displacement amount that occurring as a precursor phenomenon is difficult to
verify quantitatively. The verification data itself is impossible to acquire on the ground,
as nobody can know the location and time of a slope failure in advance. Conversely, we
can conclude that only a millimeter wave radar device capable of visualizing the subtle
surface displacement in space-time coordinates can reach such a result.
Figure 4(b) shows the results of displacement measurements taken for about 2
weeks after the occurrence of the slope failure due to the landslides. We can also detect
Fig. 4. Three-dimensional displacement map before and after slope failure.
an anomaly in which warm and cold colors are paired, in the area framed by the black
broken line. This anomaly falls somewhat below the place where the slope failure
occurred. The accumulated sediment is most likely displaced in the subsiding direction
with time in the initial stage immediately following the slope failure. We can therefore
infer that the consolidation subsidence of the accumulated sediment is measured.
3 Development of Drone-Borne Radar System
Our group is developing a set of technologies to enable the monitoring of slope terrains
such as landslide and opencast mine slopes by applying synthetic aperture processing
and InSAR analysis to data measured by a drone-mounted radar. Placement on the
drone easily secures a line-of-sight vector from multiple directions, since the antenna
moves freely. If the orbit of the platform can be flexibly designed, the radar can
precisely grasp the three-dimensional displacement of ground surfaces of complicated
terrains on its own.
High-precision information on the position of the platform is indispensable for
applying synthetic aperture processing. In this research we employed a high-precision
GNSS module for drones, a device that realizes centimeter-class positioning using RTK
technology. The synthetic aperture processing will be performed by interlocking the
raw radar data with the high-accuracy positioning information. We set the radio wave
frequency for the radar to 12.8 to 12.95 GHz (Ku-band), the values prescribed in the
provisions on “Specific Experimental Testing Authority” under the Japanese Radio
Law (the antenna power is 1 W or less in EIRP). Table 2 provides the specifications of
the radar.
Table 2. Specification of the drone-borne radar system using microwave.

Frequency 12.9 GHz
Wavelength 23.2 mm
Band width 150 MHz
Modulation FM-CW
Radar system Synthetic aperture
Antenna power 15 dBm (31.6 mW)
Antenna gain 15 dBi
Range resolution 1m
Weight (without battery) 600 g
Power consumption 5.5 Wh
A heavier mounted payload pushes up the power consumption, which shortens the
flight time of the drone. Hence, a radar with the lowest possible weight is preferred. In
this research we reduced the total weight to 600 g (excluding the power supply) by
arranging all of the RF circuits on a single board. Figure 5 shows the appearance of the
drone-borne radar device. The system consumes about 5.5 W of power and can operate
for approximately 30 min from the power of four AA batteries. Practical tests so far
conducted have verified detection up to a distance of about 1.5 km. In the future we
will begin developing algorithms for synthetic aperture processing.
4 Conclusions
Our group has developed a ground-based radar system with the aim of constructing a
technology to monitor minute displacements of mining slopes and infrastructure
buildings by remote sensing techniques. The verification test described in this paper,
along with a number of other tests repeatedly performed, have confirmed that high-
precision minute deformation measurements can be continued over an extended period
of time. Most persuasively of all, our system has clearly visualized the precursor
phenomena of an actual slope failure and the behavior of the ground surface presumed
as consolidation subsidence of sediment in space-time coordinates.
In future work we will consider a method of correcting the phase delay due to water
vapor. We also plan to develop a function to issue slope failure alerts in advance by
applying techniques to detect tertiary creep. This may make it possible to update to a
system that can objectively evaluate the risk of slope failure.
Fig. 5. Appearance and scale of the drone-borne radar.
Drone-borne radar systems have already been developed for commercial use for
collision prevention purposes. No precedence has extended to InSAR measurement,
however, so the novelty of such an approach is high. Though still in the development
stages, we hope to develop algorithms for synthetic aperture processing and InSAR
analysis and to advance to demonstration tests towards the practical stage.
Acknowledgments. The development and verification tests for the ground-based radar system
have been supported by the members of Nittetsu Mining Co., Ltd. Our work to date on the drone-
borne radar has been supported by “Coordinate Funds for Promoting AeroSpace Utilization”
from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), JAPAN.
References
Werner, C., Strozzi, T., Wiesmann, A., Wegmuller, U.: Gamma’s portable radar interferometer.
In: 13th FIG Symposium on Deformation Measurement and Analysis (2008)
Amelung, F., Galloway, D.L., Bell, J.W., Zebker, H.A., Laczniak, R.J.: Sensing the ups and
downs of Las Vegas: InSAR reveals structural control of land subsidence and aquifer-system
deformation. Geology 27(6), 483–486 (1999)
Rose, N.D., Hungr, O.: Forecasting potential rock slope failure in open pit mines using the
inverse-velocity method. Int. J. Rock Mech. Min. Sci. 44, 308–320 (2007)
Ogawa, S., Fukunaga, T., Yamagishi, S., Yamada, M., Inaba, T.: 76-GHz high-resolution radar
for automatic driving support. SEI Tech. Rev. 192 (2018)
Deguchi, T.: Deformation monitoring using space-borne and ground based radar interferometry.
NexGen Technologies for Mining and Fuel Industries, vol. I (2017)
Deguchi, T., Sugiyama, T., Kishimoto, M.: On the R&D project of drone-borne interferometric
SAR. World Mining Congress 2018 (2018a)
Deguchi, T., Sugiyama, T., Kishimoto, M.: On the development of slope monitoring technique
using millimeter wave radar system. J. MMIJ (2018b). (in Japanese)
Geotechnical Engineering Behavior
of Mudstone Formations
of Al Wadi Al Gadid Region in Egypt
Mostaf A. Yousef(&) and Ahmed T. M. Farid
Housing and Building National Research Center, HBRC, Cairo, Egypt

mostafa.abdelfadil@yahoo.com, atfarid2013@gmail.com
Abstract. More than ninety five percent of Egypt land consists of desert
regions. According to the large of the desert areas, new cities are constructed at
these desert locations. Most of the desert regions are composed of problematic
soil formations. Special geotechnical investigation studies must be carefully
performed for these problematic soil formations. Al Wadi Al Gadid region
which located in the central part of the wesern desert of Egypt is one of the
regions which composed of a problematic soil formation. This region has a low
population which needs more cities in the future. In this research, the
geotechnical engineering behavior of Al Wadi Al Gadid mudstone formations is
studied. Different Laboratory tests are performed on different undisturbed
samples as bulk samples collected from different locations in the region to
predict its geotechnical engineering parameters. Mudstone formation of the area
was founded to have a highly swelling behavior under wetting conditions
compared to its dry conditions. According to that, highly special geotechnical
studies are given to this type of formations to predict its behavior for the dif-
ferent structures established on its formations. Recommendations are given to
geotechnical engineers for highly soil investigations on its mudstone formations
to predict its characteristics and overcome its change of volume in the future for
their projects as water reached the soil under foundation level.
1 Introduction
The term mudrock is the name for claystones, mudstones, siltstones, or shales. These
mudrocks refers to the fine-grained, siliciclastic sedimentary rocks in which more than
50% of the particles are smaller than 0.06 mm in size (Blatt et al. 1980, Dick and
Shakoor 1992; Grainger 1984). Mudrocks or shales constitute about 45% to 55% of
sedimentary rock sequences; thus they are often encountered in engineering con-
struction. However, it is often difficult to obtain good quality, undisturbed samples of
mudrocks for laboratory testing because they are weak and sensitive to changes in
moisture, drilling pressure, and time (Fam et al. 2003). Therefore, information about its
behavior for most engineering projects involving mudrocks is either scarce or assumed.
One way to overcome this problem is to develop a methodology that could be used to
predict the change of the behavior of mudrock or shale under wetting conditions and
reduction of their shear strength and engineering characteristics. researches has been
done to relate durability, free swell and swelling behaviors of mudrock or shale to their
https://doi.org/10.1007/978-3-030-02032-3_13
124 M. A. Yousef and A. T. M. Farid
lithological characteristics and engineering properties (Dick 1994; Sarman et al. 1994;
Greene 2001) and still more work are needed for investigation of their behavior,
especially for swelling or shrinkage conditions. Several researchers also studied the soil
swelling potential (Holtz and Gibbs 1956; Mohamed 1978; Sridharan and Prakash
2000( where, they define the significant physical characteristics which is effect on the
swell behavior. Study of the geometric pattern of the swelling mudstone may eliminate
part of the difficulty in adequately predicting its swelling behavior as investigated by
(Johnson et al 1973; Sobolevsky and Morgachiev 1973; and Radwan et al. 1989).
In this paper, the study of Wadi Al Gadid region which is located at the centeral
part of the western desert of Egypt, where mudstone or shale formation is appeared
from surface which inter-bedded by sandstone layers in some places. This mudstone
formation of the region is varied in its clay percentages and its composition which
causes the effect of the swelling behavior. The swelling heave expected under struc-
tures established above this type of mudrock or shale formations should be predicted
according to each structure loads. Thus, geotechnical investigations of these sites
should be studied carefully for the different location in the region as indicated by
(Dhowian 1981; Dhowian 1984). Due to climatic conditions characterized by rare
rainfall, these mudstone formations are practically in dry conditions. Cracks can be
appeared in structures established directly on these formations without ant treatments,
especially for light structures due to the heave of soil as water reached in the future.
Consequently, structural problems in buildings have become more frequent and serious
if no precautions and soil treatment recommended for these type of problematic
mudstone formations in the geotechnical reports.
2 Geology of Study Area
The study area of Al Wadi Al-Gadid located within the centeral part of the western
desert of Egypt. Figures 1 and 2 show the location of the study area with respect to
Egypt map and with respect to Al Wadi Al Gadid (New Valley) region, respectively.
The geology map of the study area has been subject to many investigations and study
under the (New Valley Project) by many researchers such as: (Issawi and Jux 1982;
Said and Kerdany 1961; Kostandi 1963; and Ghobrial 1967). The Jurassic-Campanian
sequence includes the predominantly continental sandstone and clay beds that were
formely lumped under the term (Nubia). Recently, upon the detection of the marginal
marine shale or mudrock strata which occur presently in outcrops at certain strati-
graphical intervals with the sandstone strata below. The proposed study area has dif-
ferent buildings categories ranged from two floors up to six floors which represent light
to medium stress on soil. More than 32 boreholes of depths ranged between 10 m and
15 m are drilled in the study area for constructing of residential buildings, warehouses,
and roads. Seven additional test pits are executed down from the foundation surface
which is three meters to extra four meters below that depth. Undisturbed dry Mudrock
samples are extracted from these test pits for laboratory testing.
Geotechnical Engineering Behavior of Mudstone Formations 125
Fig. 1. Location of Al Wadi Al Gadid region on Egypt map
Fig. 2. Location of study area of Al Wadi Al Gadid region on map
Soil investigation performed at the study area shows that, the site consists of
medium severely weathered and crushed to closely fractured mudstone formations
which appeared from the ground surface down to the end of boring depth. Figure 3
shows sketches for the site mudrock formation at the site surface and also along the
side excavation which shows the reddish brown mudstone color and its structural
shape.
Fig. 3. Mudstone or shale shape at the site
3 Laboratory Test Program
As the site consists of mudrocks from the surface strata down to end of drilling, special
laboratory tests should be performed on this type of formation. Mudrocks of the site
consists of mudstone and claystone which could have swell behavior according to their
clay content. Tests are performed on different samples collected from the site. In this
study, mudrock behavior was investigated by different tests performed on mudrock
samples. Swelling tests according to (ASTM D-4546) is performed on mudrock
samples to predict the swelling potential and free swell ratios. Atterberg limits are
performed on the tested samples to get the activity values of the mudrock samples.
Also, hydrometer tests are performed on mudrock samples to get the percentages of
clay content samples for classification according to (Potter et al. 1980). X-ray
diffraction tests also performed for prediction of the mudrock minerals and chemical
composition.
4 Results of Laboratory Tests
Hydrometer Tests for Mudrock Samples

Clay percentages of the different types of samples are performed to classify mudrock
samples. Hydrometers tests are performed on dry fine materials passing sieve No.200.
Classification was defined according to (Potter et al. 1980) method, which define that
clay percentage is ranged between 24–36% for the tested mudrock samples. The clay
percentage is important value for predicting the mudrock activity. Activity which is
defined as the ratio of the plasticity Index (liquid limit (LL) minus plastic limit (PL)) to
percent of clay fraction finer than 0.002 mm is one means of classifying expansive soils
based on their index property. The activity index defines the capability of mudrock to
swell. According to [Skemton 1953] clays are classified with respect to their activity, as
Table 1. Classification of expansive soil according to activity

Degree of activity Activity
Inactive less than 0.75
Normal 0.75–1.25
Active greater than 1.25
Table 2. Index properties of Mudrock tested samples

Sample no. Depth (m) LL (%) PL (%) PI (%) Free swell (%) Activity
S1 4.0 48 20 28 110 1.4
S2 4.0 47 19 28 95 1.3
S3 4.5 53 19 34 125 1.6
S4 4.5 54 21 33 120 1.5
shown in Table 1. The activity of the tested samples shows that their activity index is
more than 1.25 as shown in Table 2, which describe that the mudrock formations of the
study area is in active state to swell.
Atterberg Limits Tests
Atterberg limit tests on mudrock samples are performed. The plastic limit (PL), Liquid
limit (LL) and plasticity index (PI) of the tested samples is shown in Table 2.
Free Swelling test
The free swelling test is one of the most commonly used simple tests for estimating soil
swelling potential. This test is performed by pouring 10 cm3 of dry soil, passing
through sieve no 40 (0.425 mm diameter), into a 100 cm3 graduated cylinder. The
cylinder is then filled with distilled water and the swelled volume of the soil is mea-
sured after the material settles [Egyptian code-part 2, 2001]. Where, free swelling is
then given by:
Fs ¼ ðV VoÞ=VO 100
Where:-
Fs = Free swelling, V = Final volume after swell and VO = Volume of dry soil,
10 cm3
Results of the free swelling tests are given in Table 2. The results illustrated that the
free swelling ratio of the studied samples is ranging from 95 to 125%, and that means
that the mudrock formation have the ability to swell as water reached to it in the future.
According to the change in volume of mudrock formation as water reached, there is
an expectation of superstructures defects due to the clay activity in the mudrock is
intermediate to high plasticity.
Mineralogical Composition
X-ray diffraction method was utilized for identification of the different types of clay
minerals present in the tested mudrock samples. The mineralogical analysis of
representative sample was carried out using PW1710 BASED diffractometer with a
generator operating at 40 kV, 30 mA. The mineralogy of studied mudrock samples
provides the basis for understanding their geotechnical behavior. It is also helps to
identify types of their clay minerals such as high activity smectite minerals, calcite,
quartz and other minerals. It was found that, the clay minerals present in the study area
fraction were montmorillonite as major constituent, as well as kaolinite and discrete
illite as minor constituents. The clay as montmorolonite in the soil is a good indication
of the swell potential. These problems of excessive expansive characteristics lead to
much damage to the structures built in and on these types of mudrock formation. The
test results are shown in Fig. 4 for the different minerals.
Fig. 4. X ray diffraction results for mudrock of Al Wadi Al Gadid
Chemical Analysis
Chemical analysis is carried out by X-ray fluorescence spectrometer (XRF) technique
using the instrument JEOL, JSX 3222, Japan. Chemical analysis of studied samples
indicated that they have high content of aluminum oxide (Al2O3) and ferric oxide
(Fe2O3), which are the main elements of clay minerals and where the increasing of
swelling strain of these samples related to the increasing of aluminum oxide (Al2O3)
and ferric oxide (Fe2O3), the percentage of chemical composition of (Al2O3) and
(Fe2O3) are 14.60% and 7.73%, respectively as shown in Table 3.
Swelling Pressure Tests
Swelling pressure (Sp) of a soil is the external pressure that needs to be placed over a
swelling soil to prevent volume increase. The most reliable means of measuring
Table 3. Results of chemical analysis of the mudrock

Chemical composition
SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O TiO2 P2O5 Cl- L.O.I
50.80 14.60 7.73 2.04 4.79 0.21 0.52 3.07 0.95 0.22 0.32 14.50
swelling pressure is laboratory determination using one-dimensional consolidometer.

Three different test methods are used for the determination of swelling pressures to the
site mudrock samples according to (Egyptian code- part 5, 2007). These methods are
called the direct measurement for the swelling pressure.
The first method is by using the different Pressure and this method is the closest
way to represent what happens in nature. Swelling pressure in this method is defined as
the stress required keeping the sample size of the soil stable after being fully saturated
with water under this stress. In this method, three or more identical samples are pre-
pared and each sample is loaded with different stress. The samples are submerged and
their vertical movement is monitored until the increase in volume ceases. The rela-
tionship between the different applied pressure with their heave values is drawn as
shown in Fig. 5a. By extending the obtained curves to the zero heave value as shown in
Fig. 5b, the swelling pressures of the tested mudrock samples will be obtained after
drawing the applied pressure using the logarithmic scale. According to the different
pressure method, the swelling pressure was ranging between 1020 kPa and 1250 kPa
for the S3 and S1 mudrock samples, respectively as shown in Fig. 5-b.
The second method is performed by using the pre-swelling method by applying
external pressure stresses that must be exerted on a swell mudrock sample until it
returns to its original size. In this method, one sample is installed in the Oedometer
device and submerged and allowed to swell under the influence of a small value
pressure of 10 kPa.
The vertical movement is monitored until the increase in volume ceases. Then,
external pressures are increased until the sample is compressed again and returned to its
original size and this final pressure represents the swelling pressure.
Figure 6 illustrated that the swelling pressures of the three tested mudrock samples
S2, S3 and S4, respectively. The swelling pressures using the pre-swelling method was
ranged between 950 kPa, 1220 kPa and 1400 kPa for S2, S4 and S3, respectively.
While, the third test is performed by using the method of constant volume as the
confined sample is immersed in water and no volume change is allowed. The resulting
maximum pressure at which no more stress value on immersed sample is needed with
time defined as the swelling pressure. This method is the least representative of what
happens in nature. In this method, one sample is used to control the size to remain
constant when immersed in water. The vertical movement is prevented using external
stress until the sample compressed. Computerized Oedometer apparatus was used to
maintain no volume change with time. Figure 7a illustrated a swelling pressure of
1145 kPa was achieved by using the computerized constant volume tested mudrock
sample. While, Fig. 7b illustrated that the heave was nearly neglected during the
constant volume change method.
Fig. 5. (a) Relationship between applied pressure and swell potential using different pressure
method for S1 and S3 mudrock samples (b) Swelling pressure of mudrock using the different
pressure method for S1 and S3 mudrock samples
Swell Potential Testes

Swell potential (S %) is defined as the percentage increase in the original height of the
specimen. In the swell potential test, the specimen is placed in the oedometer set under
a small pressure of nearly about 10 kN/m2, water is then added to the specimen and the
expansion of the volume of the specimen is measured until an equilibrium is reached.
The swell potential is calculated as follows:
Sð%Þ ¼ ðDHÞ=HÞ 100
Where:
DH = Height of swell due to the saturation, H = Original height of the specimen.
From the swelling pressure tests of the different pressure and pre-swelling methods
classified above, the value of the swell potential was determined while the third method
of constant volume has no swell potential as it was maintain to be nearly zero during
the test. Figures 5a and b illustrated that, the swell potential value using the different
Fig. 6. Swelling pressure using pre-swelling pressure method for S2, S3, and S4 mudrock
samples
Fig. 7. (a) Swelling pressure of mudrock using constant volume method (b) Applied pressure on
mudrock sample during the constant volume method
pressure method, was ranged between 14.5% and 16.3% for S1 and S3 mudrock
samples, respectively. While Fig. 7a, b and c illustrated the swell potential value using
pre-swelling method, was ranged between 14.1%, 15.3% and 16.2% for S2, S4 and S3,
respectively.
5 Conclusions
In our study for the mudrock or shale formations of Al Wadi Al Gadid region at the
central part of western desert of Egypt, the following conclusions are summarized:
1. Mudrock formation of Al Wadi Al Gadid has a significant swelling potential
according to its activity index which is more than 1.25 as shown in Table 2 and also
as its free swelling ratio was ranged between 95% and 125%.
2. Chemical analysis of mudrock formation indicated that it has a high content of
aluminum oxide (Al2O3) and ferric oxide (Fe2O3) where their percentages was
about 14.60% and 7.73%, respectively from the tested sample using X-ray
fluorescence spectrometer (XRF) technique. Both aluminum oxide (Al2O3) and
ferric oxide (Fe2O3) are the main elements of clay minerals which cause the
increasing of swelling strain of the mudrock formation.
3. According to the three different pressure method performed on the dry undisturbed
mudrock sample, the swelling pressure was ranging between 1020 kPa and
1250 kPa, while using the pre-swelling method it was ranged between 950 kPa and
1400 kPa. In the constant volume test method, the swelling potential was nearly
about 1145 kPa.
4. The swell potential value using the different pressure method was ranged between
14.5% and 16.3%, while the swell potential value using pre-swelling method was
ranged between 14.1% and 16.2%.
5. Combined with the consistency of the material (Atterberg limits/Activity), free
swell tests and chemical analysis, it can be concluded that in all likelihood a
Calcium Montmorillonite dominates the clay fraction.and expected to swell as water
reached to the underneath formations and cause risk to any superstructures con-
structed above it as its volume change.
6. Geotechnical engineers should perform preliminary tests such as clay content and
attereberg limits for mudrock formations of the study region of Al Wadi Al Gaid
before performing any detailed investigations tests to have knowledge about the
swell behavior of the formation.
Acknowledgments. The author appreciates Engineer Hayel Elnaggar and physical laboratory
staff of the Geotechnical Engineering Laboratory (GEL) of the Housing and Building National
Research Center (HBRC) for their help in performing the laboratory tests.
References
Awad, G.H., Ghobria, M.G.: Zonal stratigraphy of El-Kharga oasis. Geol. Surv. of Egypt, Paper
No. 34 (1965)
Blatt, H., Middleton, G.V., Murray, R.C.: Origin of Sedimentary Rocks, pp. 782–789. Prentice
Hall, Inc., Englewood Cliffs, New Jersey (1980)
Dhowian, A.W.: Characteristics of Al-Ghatt collapsing and swelling soils. In: Proceedings of
Symposium on Geotechnical Problems in Saudi Arabia, vol. 1, pp. 3–33 (1981)
Dhowian, A.W.: Characteristics of expansive soils in the Northern Region of Saudi Arabia. In:
Proceedings of the 5th International Conference on Expansive Soils. Adelaide, South
Australia (1984)
Dick, J., Shakoor, A.: Litho logic controls of mudrock durability. Q. J. Eng. Geol. 25, 31–46
(1992)
Egyptian code: Egyptian code of soil mechanics. In: Foundations Carrying Out and Designation,
Part 2, Laboratory Tests, Egypt (2001a)
Egyptian code: Egyptian code of soil mechanics. In: Foundations Carrying Out and Designation,
Part 5, Problimatic Soils, Egypt (2001b)
Fam, M., Dusseault, M., Fooks, J.: Drilling in mudrocks: rock behavior issues. J. Petrol. Sci. Eng.
38, 155–166 (2003)
Grainger, P.: The classification of mudrocks for engineering purposes. Quat. J. Eng. Geol. 17(4),
381–387 (1984)
Ghobrial, M.G.: The structural geology of the Kharga oasis. Geol. Surv. Egypt, paper 43 (1967)
Greene, B.: Predicting the unconfined compressive strength of mudrocks for design of structural
foundations. Unpublished Ph.D., Dissertation, Department of Geology, Kent State University,
Kent, Ohio, pp. 271 (2001)
Holtz, W.G., Gibbs, H.J.: Engineering properties of expansive clays. Trans. ASCE 121, 641–663
(1956)
Issawi, B., Jux, U.: Contributions to the stratigraphy of the Paleozoic Rocks in Egypt. Geology
Surv., 28–39, Egypt (1982)
Johnson, L.D., Sherman, W.C, Mcanear, C.L.: Field test sections on expansive clays. In:
Proceedings of 3rd International Conference on Expansive Soils, vol. 1, pp. 239–248. Haifa,
Israel (1973)
Kostandi, A.B.: Eocene facies maps and tectonic interpretation in the western desert, Egypt. Rev.
Inst. Franc. Pet.E. 18, 1331–1343 (1963)
Mohamed, A.A.: Engineering characteristics of swelling clays. M.Sc. Thesis, Assiut University,
Assiut, Egypt (1978)
Radwan, A.M., Abdel-Rahman, A.: Effect of macrofabric on the behavior of expansive clays. In:
Proceedings Al-Azhar Engineering 1st Conference, vol. 4, pp. 111–120. Cairo (1989)
Said, R., Kerdany, M.T.: The geology and micropaleontology of Farafra oasis. Micropaleon-
tology 7, 317–336 (1961)
Sobolevsky, Y.A., Morgachiev, V.N.: Filtration anisotropy of swelling foundation soils and
slopes. In: Proceedings of 6th International Conference on Expansive Soils, vol. 1, pp. 177–
182. New Delhi, India (1973)
Potter, P.E., Maynard, J.B., Pryor, W.A.: Sedimentology of Shale, p. 306. Springer-Verlag, New
York (1980)
Radwan, A.M., Abdel-Rahman, A.: Effect of macrofabric on the behaviour of expansive clays.
In: Proceedings Al-Azhar Engineering 1st Conference, vol. 4, pp. 111–120. Cairo (1989)
Sarman, R., Shakoor, A., Palmer, D.: A multiple regression approach to predict swelling in
mudrocks. Bull. Assoc. Eng. Geol. 31(1), 107–121 (1994)
Skempton, A.W.: The Colloidial activity of clays. In: Proceeding 3rd International Conference on
Soil Mechanics and Foundation Engineering, vol. 1, pp. 57–61. Switzerlandz (1953)
Sridharan, A., Prakash, K.: Classification procedures of expansive soils. In: Proceedings of
Institution of civil Engineers, Geotechnical Engineering, vol. 143, pp. 235–240 (2000)
Case Study of Rupture and Recovery
in Excerpt from BR-060 in the Municipality
of Alexânia in the State of Goiás, Brazil
Rideci Farias1(&), Tiago Martias Lino2(&), Haroldo Paranhos1(&),

Itamar de Sousa Bezerra3(&), Ranieri Araújo Faria Dias4(&),
Alexsandra Maiberg Hausser5(&),
and Roberto Pimentel de Sousa Júnior6(&)
1
Reforsolo Engineering, Brasília, Brazil
rideci.reforsolo@gmail.com, reforsolo@gmail.com
2
Iesplan, Brasília, Brazil
tiagoltmatias@gmail.com
3
Maccaferri, Brasília, Brazil
Itamar@maccaferri.com.br
4
University of Pará, UFPA, Belém, Brazil
ranierileislie@yahoo.com.br
5
UniCEUB, Brasília, Brazil
alexsandramaiberg@hotmail.com
6
University of Brasília, UnB, Brasília, Brazil
eng.robertopimentel@gmail.com
Abstract. The BR-060/GO rupture occurred on the border between the Federal
District and the state of Goiás, near the city of Alexânia/GO. The landslide
occurred in the early hours of December 28, 2010 at km 24 of BR-060. On
average, the highway passes about 60 thousand vehicles per day. The region is
characterized by cerrado vegetation and rugged terrain with hills, valleys and
lowlands, which contributed to this. The region is also surrounded by several
water courses. The landslide resulted in a crater about 60 meters long and 12
meters deep, completely banning road and vehicular traffic in the region. Despite
the seriousness of the accident, fortunately there were no casualties. The
emergency works began on December 28, 2010. According to the National
Department of Transportation Infrastructure (DNIT), these highway recovery
works had a 180-day deadline. The contract also included the execution of the
deviations for the liberation of vehicle traffic. The competent and supervising
body of the work was the DNIT itself, in which it gave all the support and
accompaniment to the execution of the services. In addition to the recovery of
the slopes, these works had as objective the execution of the surface and
underground drainage devices, as part of the solution. The underground drainage
was characterized by the execution of deep drains, sub-horizontal drains and
draining layers. Already the superficial drainage by the execution of gutters of
central bed, gutters of sidewalk, descents of water, tubular manhole of concrete
and boxes collector. For the recovery of the embankment, gabion containment,
paving and environmental recovery services were also carried out. Therefore,
this article aims to present analyzes of the possible causes of landslide, but also
the solutions adopted to recover the damaged section. The main solution

https://doi.org/10.1007/978-3-030-02032-3_14
136 R. Farias et al.
includes the gabion restraint system, associated with complementary works to

recover the ruptured section, but also the consequent release of traffic on the
road.
1 Introduction
This article aims to present an analysis of the possible causes of landslide, but also the
solutions adopted for the recovery of the stretch that broke from BR-060/GO on the
border between the Federal District and the state of Goiás, near the city of Alex-
ânia/GO. The rupture occurred in the early hours of December 28, 2010 at km 24 of
BR-060. On average on the highway it passes about 60 thousand vehicles per day. The
main solution for recovery of the damaged section includes the gabion restraint system,
associated with complementary works to recover the stretch for the release of traffic on
the road.
2 Location of Damaged Stretch
The section that suffered rupture is located near the border of the Federal District and
the state of Goiás, near the city of Alexânia/GO. The region is characterized by cerrado
vegetation and rugged topography with hills, valleys and lowlands, and areas subject to
runoff and water courses. Figure 1 shows the general location map of the area where
the highway rupture occurred, which is approximately 15 km from the city of
Alexânia/GO.
Fig. 1. Localization of BR 060/GO disruption (DNIT 2011).
The landslide resulted in a crater near 60 meters in length and 12 meters in depth,
completely banning the route and traffic of vehicles in the region. The intense rainy
season would be the main cause of the landslide, and fortunately no casualties, despite
the seriousness of the accident. Figures 2, 3, 4, 5 and 6 show the section that broke.
Due to what happened there was a need to divert the traffic in two alternatives to the
driver who left the Federal District to Goiânia. One by BR-070, towards Cocalzinho, to
GO-414 and then to BR-060, which increases the route to the capital of Goiás by
28 km. The other option was to reach the GO-010 by Luziânia and continue until
Case Study of Rupture and Recovery in Excerpt from BR-060 137
Fig. 2. Broken section of BR-060/GO (12/31/2010).
Fig. 3. Broken section of the BR-060/GO (12/31/2010).

Fig. 7. Alternative paths when the BR-060 interdiction.
Goiânia. In this case, it is 50 km more, bringing the distance between Brasilia and
Goiânia to 250 km. Figure 7 shows alternative routes for motorists when the BR-060
interdiction.
3 Analysis of Possible Causes
3.1 High Pluviometric Index - Soil Saturation

The rainy period in the region, with great intensity between the months of December
and January, increased soil saturation, leading to erosive processes, mainly internal
erosion, but also an increase in the value of percolation and the consequent decrease of
soil resistance at sites where the flow of water was more concentrated. In addition,
upstream of the cut made to install the highway, there is a considerable area in which
the native vegetation was suppressed, aiming investments in agriculture and livestock.
In addition to the suppression of vegetation, it was verified that the basin was placed in
the ground for water storage, which, when reconciled with the infiltration and runoff,
caused the possible preferential flow path in the soil mass to increase the saturation of
the subsoil to the highway. Thus, it was pointed out as one of the main causes of the
rupture of the road, the degree of saturation of the soil and the increase of the pore
pressure, caused by the rains, with the consequent destabilization of the massif.
Figure 8 shows the possible flow inside the bulkhead, and Fig. 9 shows a view of the
area with the infiltration basin upstream of the stretch of the highway that has
undergone the rupture.
Fig. 8. Possible flow inside the massif (Lino 2016).
Fig. 9. Infiltration basin upstream of the highway (Lino 2016).
Fig. 10. View of saturated soil after landslide, December 2010 (Lino 2016).
Figure 10 shows the saturation inside the massif that with the excavations for soil
removal, the flow of water was visible.
Despite some data and initial characteristics of the work, one could not say what
had really caused the highway breakup. For this, more detailed investigations of the
region were carried out, as follows.
3.2 Geotechnical Investigation

In addition to inspections in the area of the accident, it was also carried out for the
execution of a SPT type survey in order to verify the conditions of the subsoil,
groundwater of the region, besides assisting in the basic design studies.
At first there were three drill holes totaling 45.35 (forty-five meters and thirty-five
centimeters). These probes were named SPT 01, SPT 02 and SPT 03, as shown in
Fig. 11, with SPT 1 at the edge of the highway and SPT 02 and SPT 03 on the axis of
each of the tracks.
Fig. 11. Location of SPT drill holes (BR-060, km 24).
Fig. 12. Execution view of the SPT 1 survey (BR-060, km 24).
Figure 12 shows the execution of the SPT 01 probe at the edge of the highway.
When considering the type of soil found, it was found to be predominantly a layer
of landfill + material of the estradiol body with a variable thickness of up to three
meters. Underlying this layer, clayey to slightly clayey/sandy to slightly sandy, varying
up to 3.0 meters thick. Then, clay-silt/sandy clay, little clay. In terms of water level the
SPT 01 probe was 1.50 meters from the “mouth” of the hole. In SPT 02 at 6.10 meters,
and at SPT 03 at the depth of 7.15 meters. In view of this, it was possible to establish
the possible behavior of the groundwater table in the massif.
4 Stability Analysis of Slope
For the proper analysis of slope stability, the on-the-spot checks were added to cor-
relations with the SPT probes to choose the parameters.
In the simulation with the unsaturated soil, 10 kPa was used for cohesion, 30° for
the friction angle and 17 kN/m3 for the specific soil weight, obtained from correlations.
Figure 13 shows the slope stability safety factor of 1.454. Simulation done with the
empirical data of the dry soil, value according to the minimum criteria of safety factors
of NBR 11682/2009.
Fig. 13. Stability analysis with natural soil. (FS = 1.454).
In the second simulation of the slope stability, we have as a characteristic the

calculation of the safety factor with the saturated soil. Also the correlated SPT data with
empirical tables. For saturated soil, we worked with 0 (zero) kPa for cohesion,
remaining 30° for the friction angle, although it was lower than the safety factor, and
17 kN/m3 for the specific soil weight. Figure 14 shows the safety factor of 1.117. Note
the expected reduction of the safety factor with the cohesion of 0 kPa, considering the
saturated soil. Although the safety factor is greater than 1, it is not in accordance with
the safety criteria of NBR 11682/2009.
Fig. 14. Stability analysis with saturated soil. (FS = 1.117).

5 Solutions for Recovery of the Damaged Area
The project developed by DNIT for the emergency execution of the work involved a
series of solutions with surface and underground drainage, containment with gabion,
paving, but also the environmental recovery of the area affected by the works.
5.1 Surface and Underground Drainage

The internal or subterranean drainage aims to remove the percolation water from the
massif (flow in the soil pores), reducing percolation flow. In the progress of excavation
and removal of saturated soil, one of the concerns of the construction company was to
channel the flow of water to an appropriate destination. Deep drainage systems were
used (deep drain with geotextile blanket, sub-horizontal drain, DHP, and drainage
mattress). Figures 15, 16, 17, 18, 19 and 20 show part of the execution of the deep
drainage. Figure 21 shows part of the surface drainage with the finished work.
Fig. 15. Excavations for the installation of deep drains.
Fig. 16. Execution of deep drains.
5.2 Containment in Gabon

A widely used solution, whether for landfill stability, containment, surface protection of
slopes, water courses and highways. They are elements of prismatic or cylindrical form,
constituted by a metallic net and filled with stones of hand or rolled pebbles, destined to
Fig. 17. Execution of deep drains.
Fig. 18. Drilling for DHP installation.
Fig. 19. Execution of the draining mattress.
the solution of geotechnical, hydraulic problems and of erosion control. Gabions can
be, (gabion type box, gabion type Reno mattresses or gabion in bags). Gabions are able
to drain the waters from percolation of the terrain and to resist all kinds of efforts, and
above all, to work under traction thus eliminating one of the main factors of instability.
To recover the landfill and to solve drainage problems, gabion structures were used.
Four gabion structures were executed, three box type and one Reno type. Figures 22
and 23 show the typical section of the highway and then in Figs. 24, 25, 26 and 27
show the execution and some of the systems in gabion, and in Figs. 28 and 29 the
finished section.
Fig. 20. Drainage mattress construction.
Fig. 21. Smooth waterfall running along the banks of the stretch recovered on Highway BR
060.
Fig. 22. BR 060 cross section in the ruptured section (DNIT 2011) (Geo-slope International
2012).
Fig. 23. Type section for the section retrieved from BR 060 (DNIT 2011) (Geo-slope
International 2012).
Fig. 24. Execution of gabion restraint.
Fig. 25. Execution of gabion restraint.
Fig. 26. View of gabion restraint.

Fig. 27. System implanted with the completion of the work.
Fig. 28. Retrieved and completed section.
Fig. 29. Retrieved and completed section.
6 Conclusions/Recommendations
This article sought to present analyzes of the possible causes of landslide, but also the
solutions adopted for the recovery of the stretch that broke from BR-060/GO on the
border between the Federal District and the state of Goiás, near the city of Alex-
ânia/GO. Thus, in the face of field studies and verifications, but also of due analysis, it
is possible to conclude:
(a) The rainy season in the region, with great intensity between the months of
December and January, increased the saturation of the soil, and the percolation
with preferential paths of flow coming to cause erosive processes, mainly internal
erosion, but also the consequent decrease of the resistance of the massif in places
where the water flow regime was more concentrated;
(b) Thus, one of the main causes of road rupture was the degree of soil saturation and
the increase in pore pressure caused by the rains, with the consequent destabi-
lization of the massif;
(c) Upstream of the cut made for the installation of the highway, there was a con-
siderable area in which the native vegetation was suppressed, aiming investments
in agriculture and livestock. In addition to the suppression of vegetation, it was
verified the installation of basin in the land for water storage that reconciled to the
infiltration and to the runoff caused the possible preferential path of flow in the soil
mass to increase the saturation of the subsoil to the highway;
(d) When analyzing possible causes of road rupture, the geological-geotechnical,
environmental and hydrological studies of a region for the execution of engi-
neering works are of great importance. With these studies one can determine the
best form and the most appropriate type of work to be implemented;
(e) Interference and human action has caused negative aspects in the environment,
consequently interfering in engineering works. These actions cause soil degra-
dation, severe erosive processes and changes in the surface and underground
dynamics of regions. In the case study in question, this interference caused the
emergence of springs and infiltration basins, factors that contributed to the
breakdown of the highway;
(f) With the analysis of the environmental problems found in the region, it is justified
the execution of the drainage devices and the containment in gabions. Under-
ground and surface drainage aims to conduct the water to suitable locations,
avoiding damage to the implanted work. On the other hand, the gabion con-
tainment structures have as main objective to contain the landfill and avoid ero-
sive processes;
(g) The importance of preventive measures instead of corrective ones in order to
avoid damages in works and (or) disturbances to the population and damage to the
environment;
(h) In order to verify the current conditions of the implanted system, the site where
the rupture occurred was visited in July 2017 and it was verified that the implanted
works are consistent with the expected operation.
Acknowledgment. The National Department of Transport Infrastructure (DNIT), JM Ter-

raplenagem e Construções Ltda., Reforsolo Engenharia Ltda., Universidade Católica de Brasília
(UCB), University Center of Brasília (UniCEUB) and the Institute of Higher Education Planalto
with important contributions that made possible the accomplishment of this work.
References
ABNT NBR 11682:2009 - Estabilidade de encostas
Departamento Nacional de Infraestrutura de Transportes. DNIT: 2011. Projeto Executivo de Obra
Emergencial para a Rodovia BR-060/GO, Div. DF/GO Entrº BR-364 (Contorno de Jataí),
Entrº GO-425 – Entrº GO-139 (Início Travessia Urbana Alexânia)
Geo-slope International: Stress deformation modeling with SIGMA/W, an engineering method-
ology, 335 p. (2012)
Lino, T.M.: Estabilização de Taludes: Trabalho de Conclusão de Especialização, Publicação
ENC. TCC– 020.1, Departamento de Engenharia Civil, Instituto de Ensino Superior Planalto,
Brasília, DF, 82 p. (2016)
Use of the Pedological Map in the Geotechnical
Characterization of Aris Mestre D’Armas
in Planaltina - DF
Haroldo Paranhos1(&), Rideci Farias1(&),

Joyce Maya Lucas Silva2(&), Paulo Sergio Pereira3(&),
Roberto Pimentel de Sousa Júnior4(&),
and Alexsandra Maiberg Hausser5(&)
1
Reforsolo Engineering, Brasília, Brazil
reforsolo@gmail.com, rideci.reforsolo@gmail.com
2
CONCREMAT, Brasília, Brazil
joyce.civil@gmail.com
3
UCB, Brasília, Brazil
psergio@ucb.br
4
University of Brasília, UnB, Brasília, Brazil
eng.robertopimentel@gmail.com
5
UniCEUB, Brasília, Brazil
alexsandramaiberg@hotmail.com
Abstract. The present work presents the results of the geotechnical studies
realized at ARIS Mestre D’Armas in Planaltina, DF, aiming to do a geotechnical
characterization of soils and geotechnical risk, using field tests, SPT tests, ring
penetrometer probes and sampling probes, permeability tests in drill holes and
soil characterization tests in the laboratory. As an auxiliary tool in the spatial-
ization and definition of the tested points of the “Geotechnical Field Investi-
gations” program, through the prediction of local soil characteristics, the
pedological and geological maps were used for the incursion of the field teams
to the main groups of soils indicated in these maps. Parallel studies were exe-
cuted in the pedogenetic slopes of the area for the evaluation of toposequence.
The overlapping of the geological and pedological maps of the area defined the
units with similar geotechnical characteristics. The methodology adopted con-
sists of the evaluation of the characteristics of the soils of the area, based on the
interpretation and descriptions contained in the pedological and geological
maps; defining a map of unconsolidated material of geotechnical interest, of the
area in question. The evaluation includes the superposition of the geological
maps and the pedological map, based on methodologies, in which adaptations of
geological and pedological classifications are made for geotechnical purposes.
Characterization and geotechnical parameters were defined from field and lab-
oratory tests on deformed and undisturbed samples, taken from strategically
chosen points, in order to confer a certain representativity of the area under
study. A correlation of the soils pedologically defined, with the geotechnical
classifications were made by considering the properties, such as: granulometry,
mineralogy, Atterberg limits, etc. and use of morphological data such as hori-
zons thickness, macrofactory, color, etc., added to the physical properties of
resistance and permeability determined in each profile. The verification of the

https://doi.org/10.1007/978-3-030-02032-3_15
150 H. Paranhos et al.
geotechnical properties of each horizon of interest, through tests was associated

to the pedological type, generating the geotechnical map of the area in question.
1 Introduction
This paper presents the results of the geotechnical studies executed at ARIS MESTRE
D’ARMAS in Planaltina/DF, with the aim of characterizing the soils by means of field
tests, SPT tests, soil tests, permeability tests in drill hole and works office. The studies
consisted in the execution of a program of “Geotechnical Field Investigations”, spa-
tialized in the area of interest, according to the main groups of soils indicated in the
pedological and geological maps.
2 Scope of Work
Initially the works were conducted in the office, with the evaluation of the existing
documents of the area in question.
Subsequently, using the geological and pedological maps for the evaluation of
toposequence, suggested and proposed by Soares et al. (1995), were carried out with
the possession of pedological maps (Scale 1: 400,000) and geological characteristics of
the region (scale 1: 250,000).
Based on the above mentioned studies, a field trip was carried out to for execute the
SPT, penetrometric and drill holes for the evaluation of the masses.
In the following, brief considerations are made on the types of soils and also on
their resistance and permeability.
3 Location of the Study Area
Figures 1 and 2 below, taken from Google Earth on July 10, 2013, show approximate
macro views of the geotechnical survey and survey points, comprising 2625.00 ha.
Fig. 1. Macro view of the area of execution of geotechnical studies in Planaltina - DF.
Use of the Pedological Map in the Geotechnical Characterization 151
Fig. 2. Macro view of the area and location of the points studied.
4 The Use of Pedology in Geotechnics
Several other interpretations can be made from soil surveys, in addition to agricultural
purposes, Larach (1983) cited by Resende et al. (1995). Baize et al. (1990) affirm that
with the creation of a pedological referential, specialists from other scientific areas will
be able to construct another typology, corresponding to their own needs, result of the
horizons of reference.
4.1 Applied Pedology

Many works are cited by Paranhos (1998) correlating several attributes of pedology
with geotechnical engineering. The overlap of geological and pedological maps of a
given area can define units with geotechnical characteristics.
4.2 Pedogenesis
The pedology, based on models, indicates five main soil-forming factors and their
properties are caused by climate, organisms, source material, time and topogaphy.
4.3 Soil Landscape

The landscape is the image of the combined action of soil formation factors, such as
topography, organisms, origin material, climate, over time. It is very important to know
the distribution of the soils in the landscape in the execution of soil (or pedological)
surveys, and also in studies of soil genesis (Prado 2005).
4.4 Local Soils

Figure 3 shows the soil group of the region under study and Fig. 4 shows the geo-
logical formation of the region.
Generally, at the top of the geopedological slope, there is the presence of more
mature soils (Dark Red Latosols and Yellow Reds), which have a high drainage
Fig. 3. Soil groups of the region (Modified - EMBRAPA, 2004).
Fig. 4. Geological Formation of the region (Modified - EEZ (2002).

capacity. The cambisols arise in specific conditions of the slope, associated to the
inclinations and the lack of the action of the weathering.
It is also observed the presence of soils that had in their formation process the
conditions of excess water and low hydraulic gradient (hydromorphism). There are also
dark plastic soils with a high content of organic matter, typical of regions of accu-
mulation of vegetal sediments.
5 Methodology
The methodology consists of the evaluation of the soil characteristics, from the
interpretation and superposition of the geological and pedological maps. This
methodology is based on works described and presented by Dias and Milititsky (1990),
Bastos (1991) cited by Dias and Milititsky (1994) and Paranhos (1998), in which
adaptations of geological and pedological classifications for geotechnical purposes are
made.
In general, the technical bulletins that accompany the pedological maps provide
important information for the different soil classes, taking into account the following
criteria: B horizon type, base saturation percentage, clay activity, type of soil horizon
A, textural class and drainage class, besides the phase criterion, taking into account the
vegetation, topography, stoniness and presence of lateritic concretions. This bulletin
also contains a detailed description of the main morphological characteristics (color,
texture, structure, porosity, etc.).
Characterization and geotechnical parameters can be defined from field and labo-
ratory tests in samples representative of the area under study.
Aerial photographs, geological maps, topographic maps and other related works
can be used to better detail the areas. This type of investigation, carried out in an
“office”, serves to avoid overlapping fieldwork, making the visits exhaustive; (geo-
logical and geological mapping), which is the fundamental idea of the proposal that
involves the present work: an interpretation of the behavior of the soil in an indirect
way, based on already existing works (geological and pedological maps). The estimate
of geotechnical units will assume initially the symbology of work published by
Camargo et al. (1987), also cited by Dias (1995), to classify the surface horizons and to
the subsurface classification, that is, horizons that keep the geological structure, will be
adopted the simplified geological classification, based on the work of IAEG (1979).
The following symbols for the geotechnical classification were adopted:
ABxyz ðiÞ
Which:
– The capital letters for superficial horizons.
– Lowercase letters for subsurface horizons.
6 Obtaining Geotechnical Characteristics of Aris Mestre

D’Armas Surface Soils, in Planaltina-Df, Using Pedological
Survey
The geotechnical characteristics of the soils under study (Fig. 5) were estimated using
the overlapping of the EMBRAPA (2004) pedological maps, in a scale of 1: 400,000
and geological ZEE (2002) in a scale of 1: 250,000, to facilitate fieldwork; rationally
defining sample collection points and polls in general.
Fig. 5. Geotechnical Units of the region and location of the points studied.
6.1 Surveys and Field Tests

The field tests were performed in order to determine the soil resistance and permeability
characteristics. The following tests were used: 14 probes, 14 probes with the ring
penetrometer and 14 permeability tests.
6.2 Evaluation and Description of Study Areas

By means of the adopted methodology, field and laboratory tests, the geotechnical map
in question (Fig. 5) and the appropriate descriptions of the geotechnical units are
suggested.
Table 1 shows the soil variation found in the region by means of the overlapping of
the pedological and geological maps, in order to generalize the geotechnical units.
Table 1. Pedological, geological and geotechnical units.

Pedological unit Geological unit Geotechnical unit
LV TQdl LVTQd
LV Mpc LVMpc
LV QHa LVQHa
LA TQdl LATQd
LA Mpc LAMpc
CX TQdl CXTQd
CX Mpc CXMpc
CX QHa CXQHa
GX TQdl GXTQd
GX QHa GXQHa
6.3 Description of the Soils

The following is a description of the proposed geotechnical units.
LVTQd - Red Latosol, Detritus-Lateritic Cover
They are very deep soils, where the thickness of horizons A and B together, is greater
than 3 meters, clayey or medium texture and rich in sesquioxides. They are very
porous, quite permeable and markedly drained.
The susceptibility to erosion varies from practically null to light.
LVMpc - Red Latosol, Paracatu Formation
Idem to the LVTQd, but presenting smaller depths in the B horizon, due to the
presence of the rock in subsurface. It is a transition zone between the Oxisols and the
cambisols.
LVQHa - Red Latosol, Holocene Alluvial
Idem to the LVTQd but presenting small depths in the B horizon. It is a transition zone
between the Latosols in the lowland regions.
LATQd - Red-Yellow Latosol, Detritus-Lateritic Cover

They are soils with high permeability of water. Very Deep Chemically, more than 95%
of the lattosoils are dystrophic and acidic, with low to medium cation exchange
capacity and pH levels around 4.0 and 5.5.
LAMpc - Red-Yellow Latosol, Paracatu Formation
Idem to the LATQd but presenting smaller depths in the horizon B, by the presence of
the rock in subsurface. It is a transition zone between the Oxisols and the cambisols.
CXMpc - Cambisol, formation Paracatu
They are soils that have subsurface horizon with little physical and chemical alteration.
It presents low degree of weathering. Generally, they are associated with more
mountainous topographies (wavy and strong-wavy). They range from shallow and
deep, reaching between 0.2 and 1 m.
CXTQd - Cambissolo, Detrito-Laterítica Cover
Idem to the CXMpc, but presenting some depths in the horizon B, by the presence of
soil on the subsuperficial rock. It is a transition zone between the Oxisols and the
cambisols in lowland regions.
CXQHa - Cambisol, Holocene Floods
Idem to the CXMpc, but presenting some depths in the horizon B, by the presence of
soil on the rock in subsurface. It is a transition zone between the Cambisols and the
Gleissolos in lowland regions.
GXQHa - Gleissoil, Holocene Mudslides
They are hydromorphic soils that generally occupy the depressions of the landscape,
subject to floods. They present drainage of the types: poorly drained or very poorly
drained, occurring frequently, thick dark layer of organic matter poorly decomposed on
a grayish (gleaned) layer, resulting from oxidation environment.
GXTQd - Gleissolo, Detrito-Laterítica Cover
Same as GXQHa, but presenting some soil properties in the unsaturated zone. Capil-
lary height generated by the presence of NA, does not reach this superficial layer. It is a
transition zone between the Oxisoils and the Gleissoils in lowland regions.
6.4 Summary of Geotechnical Characteristics

Table 2 shows the classification of the soils of the geotechnical units.
Table 3 presents the classification for resistance.
Table 4 presents the permeability classification.
6.5 Susceptibility to Erosion

According to Camapum de Carvalho et al. (1993), the occurrence of erosions in the
Federal District is related to two main erosive agents: man and water action, which
generally act associated, although in different proportions and conditioned by soil
erodibility.
Erosive Susceptibility in Studied Soils
Table 2. Soil classification of geotechnical units (SUCS, TBR and MCT).

Tipo de solos Prof. (m) SUCS TBR MCT
LVTQd 1,0 CL e MH A-7-5 LA’-LG’
3,0 CL e MH A-7-5 LA’-LG’
5,0 MH A-7-5 e A-7-6 LG’
7,0 MH A-7-5 LG’
LVMpc 1,0 CL e MH A-7-5 LA’-LG’
3,0 ML A-4 NS’
5,0 ML A-4 NS’
LVQHa 1,0 CL e MH A-7-5 LA’-LG’
3,0 SM A-2-4 NA’
5,0 SM-SC A-3 NA
LATQd 1,0 CL e MH A-7-5 LA’-LG’
3,0 CL e MH A-7-5 LA’-LG’
5,0 MH A-7-5 e A-7-6 LG’
7,0 MH A-7-5 LG’
LAMpc 1,0 CL e MH A-7-5 LA’-LG’
3,0 ML A-4 NS’
5,0 ML A-4 NS’
CXMpc 1,0 ML A-4 NS’
3,0 ML A-4 NS’
CXTQd 1,0 CL e MH A-7-5 LA’-LG’
3,0 ML A-4 NS’
5,0 ML A-4 NS’
CXQHa 1,0 SM-SC A-2-4 NA’
3,0 ML A-2-4 NA’
5,0 ML A-3 NA
GXTQd 1,0 CL e MH A-7-5 LA’-LG’
3,0 SM-SC A-2-4 NA’
GXQHa 1,0 ML A-4 NA’
3,0 SM-SC A-2-4 NA’
6.6 Red Latosols

The high velocity of water infiltration in these soils, avoids the formation of large
volumes of surface runoff and associated with the low slope of the area, provides less
susceptibility to erosion.
6.7 Yellow Red Latosols

This soil class is commonly found in water dividers and in transition areas for plateau
ridges with flat to smooth corrugated topography, associated with a secondary field of
soil with good drainage and permeability, not observing the development of processes
erosive.
Table 3. Soil classification of the Geotechnical Units (SPT and Admissible Voltage).
Tipo de solos Prof. (m) SPT Tensao admisslvel Designafao
(kgf/cm2)
LVTQd 1,0 2 a 5 0,4 a 1,2 Mole
3,0 2 a 5 0,4 a 1,2 Mole
5,0 6 a 10 1,4 a 2,2 Media
7,0 6 a 10 1,4 a 2,2 Media
LVmpc 1,0 2 a 5 0,4 a 1,2 Mole
3,0 6 a 10 1,4 a 2,2 Media
5,0 6 a 10 1,4 a 2,2 Media
LVQHa 1,0 2 a 5 0,4 a 1,2 Mole
3,0 6 a 10 1,4 a 2,2 Media
5,0 6 a 10 1,4 a 2,2 Media
LATQd 1,0 2 a 5 0,4 a 1,2 Mole
3,0 2 a 5 0,4 a 1,2 Mole
5,0 6 a 10 1,4 a 2,2 Media
7,0 6 a 10 1,4 a 2,2 Media
LAMpc 1,0 2 a 5 0,4 a 1,2 Mole
3,0 6 a 10 1,4 a 2,2 Media
5,0 6 a 10 1,4 a 2,2 Media
CXMpc 1,0 5 a 8 1,2 a 1,8 Pouco Compacta
3,0 9 a 18 2,0 a 3,2 Median. compacta
CXTQd 1,0 2 a 5 0,4 a 1,2 Mole
3,0 5 a 8 1,2 a 1,8 Pouco Compacta
5,0 9 a 18 2,0 a 3,2 Median. compacta
CXQHa 1,0 5 a 8 1,2 a 1,8 Pouco Compacta
GXTQd 1,0 0 a 2 < 0,4 Muito Mole
3,0 2 a 5 0,4 a 1,2 Mole
GXQHa 1,0 0 a 2 < 0,4 Muito Mole
6.8 Switching
These soils occur mainly on slopes and slopes with higher slopes. The high silt content
of the A horizon and the restricted depth of the profile make this kind of soil perme-
ability difficult. The combination of these features with the high slope rates where these
soils occur makes them more susceptible to erosion.
6.9 Gleissoil
The limitations and the environmental sensitivity of these soils are very large, so much
that they are considered as an area of environmental preservation. The limitations for
Table 4. Classification of soils of geotechnical units regarding permeability.

Tipo de solos Prof. (m) Ordem de Designação
permeabilidade
LVTQd 1,0 10−4 a 10−2 Permeável
2,0 10−4 a 10−2 Permeável
LVMpc 1,0 10−4 a 10−2 Permeável
LVQHa 1,0 10−4 a 10−2 Permeável
LATQd 1,0 10−4 a 10−2 Permeável
LAMpc 1,0 10−4 a 10−2 Permeável
2,0 10−6 a 10−4 Median. Permeável
CXMpc 1,0 10−6 a 10−4 Median. Permeável
CXTQd 1,0 10−4 a 10−2 Permeável
CXQHa 1,0 10−4 a 10−2 Permeável
GXTQd 1,0 10−4 a 10−2 Permeável
2,0 10-8 a 10−6 Pouco Permeável
GXQHa 1,0 10−4 a 10−2 Permeável
2,0 10-8 a 10−6 Pouco Permeável
Table 5 Classification of soils of geotechnical units regarding erosive susceptibility.

Tipo de solos Classe de Erodibilidade
LVTQd Muito Baixa a baixa
LVMpc Muito Baixa a baixa
LVQHa Muito Baixa a baixa
LATQd Muito Baixa a baixa
LAMpc Muito Baixa a baixa
CXMpc Forte
CXTQd Forte
CXQHa Forte
GXTQd Muito Baixa a baixa
GXQHa Muito Baixa a baixa
urban use are due to its high plasticity, common presence of organic matter and low
permeability. Table 5 presents the summary of the erodibility class of the soils of the
area.
6.10 Susceptibility to Flooding

For the area in question, the risk of flooding is only likely to occur in the regions
delimited by soils classified as Gleissoils. These soils demarcate the strip with direct
influence of streams and adjacencies.
6.11 Susceptibility to Colapse

The evaluation of soil collapse was done indirectly through the interpretation of SPT
surveys in the area and the soil profile. In general, the collapsible soils are soils with the
soft macroporous structure, with high porosity or high void index. They are usually
unsaturated soils and have a low SPT in the assessment altitudes.
7 Conclusions
The generation of an unconsolidated material map, based on the overlapping of

Pedological and Geological maps, for characterization of areas is a tool of great
geotechnical potential, although the limit positions of the units and the transitions
between them should be better evaluated by field and located probes.
The limitations of the present study are also highlighted, since databases were used
at regional scales (1: 400,000 and geological scale on the 1: 250,000 scale).
References
Baize, D., Girard, M., Boulaine, J., Cheverry, C.L, Ruellan, A.: Um Referencial Pedológico.
Porquê? (1990)
Dias, R.D., Militissky, J.: Metodologia de Classificação de Unidades e Perfis Geotécnicos
Desenvolvida na UFRGS. SOLOS E ROCHA 17, 81–92 (1994)
Dias, R.D.: Proposta de Metodologia de Definição de Carta Geotécnica Básica em Regiões
Tropicais e Subtropicais, Revista do instituto geológico, volume especial, pp. 51–55 (1995)
Embrapa Cerados: Mapa pedológico digital – SIG atualizado do Distrito Federal escala 1.100.00
e uma síntese do texto explicativo, 31 p. (2004)
IAEG: Association for Engineering Geology and the Environment (1979)
Paranhos, H.S.: Caracterização Geotécnica dos Principais Grupos de Solos do Mapa Pedológico
do Distrito Federal (Esc. 1:100.000): Estudos na Área de Dinamização Urbana, entre
Samambaia e Gama (1998)
Prado, H.: Pedologia Fácil - Aplicações em solos tropicais (2005)
Resende, M., Curi, N., Rezende, S.B., Corrêa, G.F.: Pedologia: base para distinção de ambientes,
Jard, Neput, Viçosa, 304 p. (1995)
Soares, A.M.L., et al.: Áreas degradadas suscetíveis aos processos de desertificação no Ceará.
IPEA, Brasília (1995)
ZEE: Projeto Zoneamento Ecológico-Econômico da região integrada de desenvolvimento do
Distrito Federal e Entorno – Mapa Geológico – Fase I (2002)
Author Index
A K
Ai, Fei, 94 Khachay, Andrey, 31
Al Mandalawi, Maged, 10 Khachay, Oleg, 31
Al-Arifi, Nassir S. N., 60 Kishimoto, Munemaru, 115
Anbazhagan, P., 60 Kusumayudha, Sari Bahagiarti, 83
B
L
Bajaj, Ketan, 60
Lino, Tiago Martias, 135
Balachandar, Ram, 73
Liu, Jian, 94
C
Chen, Pengfei, 94 M
Choiriyah, Siti Umiyatun, 83 Maji, Soumen, 73
Ciptahening, Ayu Narwastu, 49 Mollamahmutoğlu, Murat, 39
Correia, António Gomes, 105 Moustafa, Sayed S. R., 60
Cortez, Paulo, 105 Murwanto, Helmy, 83
D N
Dahlhaus, Peter, 10 Nugroho, Nandra Eko, 49
de Sousa Bezerra, Itamar, 135
de Sousa Júnior, Roberto Pimentel, 135, 149 P
Deguchi, Tomonori, 115 Paranhos, Haroldo, 135, 149
Dias, Ranieri Araújo Faria, 135 Peng, Wanlin, 94
Dowling, Kim, 10 Pereira, Paulo Sergio, 149
Dwivedi, Ajay, 1 Phienwej, Noppadol, 49
F R
Farias, Rideci, 135, 149 Roussinova, Vesselina, 73
Farid, Ahmed T. M., 123
S
H Sabry, Mohannad, 10
Hachay, Olga, 31 Samanta, Manojit, 1
Hanmaiahgari, Prashanth Reddy, 73 Sarkar, Shantanu, 1
Hausser, Alexsandra Maiberg, 135, 149 Sharma, Mahesh, 1

https://doi.org/10.1007/978-3-030-02032-3
162 Author Index
Silva, Joyce Maya Lucas, 149 Y

Sugiyama, Tomoyuki, 115 Yan, Xiuping, 94
Sutarto, 83 You, Greg, 10
Yousef, Mostaf A., 123
T
Tinoco, Joaquim, 105 Z
Toll, David, 105 Zhou, Fan, 94

Recent Research On Engineering Geology and Geological Engineering

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Recent Research On Engineering Geology and Geological Engineering

Загружено:

Авторское право:

Доступные форматы

Sustainable Civil Infrastructures

More information about this series at http://www.springer.com/series/15140

ISSN 2366-3405 ISSN 2366-3413 (electronic)

Library of Congress Control Number: 2018957283

© Springer Nature Switzerland AG 2019

Slope Stability Analysis and Suggestive Measures for an Active

Volcanic Disaster and the Decline of Mataram Kingdom in the Central

Dr. Janusz Wasowski

Dr. Janusz Wasowski is a research geologist at

Since 2007, he has been a member of the Editorial

Dr. Tom Dijkstra

Shantanu Sarkar(&), Manojit Samanta, Mahesh Sharma,

Geotechnical Division, CSIR-CBRI, Roorkee, Uttrakhand, India

Abstract. Landslides are frequently occurring phenomena in Indian Himalayas

© Springer Nature Switzerland AG 2019

Fig. 2. A panoramic view of the landslide

Fig. 3. Highway at risk due to debris slide

Normal Stress (kg/cm2)

Fig. 4. Shear envelope of sample obtained from direct shear test

Table 1. Physical properties of soil samples

Table 2. Results of shear test

4 Slope Stability Analysis and Mitigation Measures

Fig. 5. Sections for stability analysis

Fig. 6. Slope stability analysis without considering remedial measures

Table 3. Factor of safety of MN and CD sections

Table 4. Speciﬁcation of soil nails

Acknowledgments. Authors are grateful to the Director, CSIR-Central Building Research

Maged Al Mandalawi1(&), Greg You1, Peter Dahlhaus1,

Keywords: Circular failure Toppling failure Factor of safety

© Springer Nature Switzerland AG 2019

2 Site Conditions and Engineering Geology

Fig. 5. Highly fractured nature of the Magazine Shale

3 Description of the Failure

slope went through a combination of circular–toppling failure mechanism, as shown in

4 Back Analyses: Methods and Results

4.1 Probabilistic Back–Analysis of Material Strength

4.2 Back–Analysis of Groundwater Conditions

does not correspond to the groundwater levels observed in previous hydrological

4.3 Toppling Limit Equilibrium Back–Analysis

4.4 Back–Analysis Using Finite Element Method (FEM)

4.5 Kinematic Analyses

5 Discussion of the Back Analysis

Fig. 19. Kinematic stability analyses of area around the failure

Olga Hachay1(&), Andrey Khachay2, and Oleg Khachay2

Abstract. A new method for modeling acoustic monitoring of a layered-block

Keywords: Hierarchical environment Acoustic ﬁeld Iterative algorithm

© Springer Nature Switzerland AG 2019

1.1 Algorithm of Modeling of Sound Diffraction on 2D Composite

We calculate ul ðM 0 Þ; M 0 2 Pj in the layer where the second hierarchical elastic

rðj þ 1Þim ðk1ðj ZZ

We calculate um ðM 0 Þ; M 0 62 S2Cm ; 2 Pj þ 1 in the layer where the third hierarchical

Let: l = l + 1; m = m + 1; s = s + 1. If l < 3 or l = 3, then we turn to the algorithm

u0m1 ðM 0 Þ ¼ us1 ðM 0 Þ; M 0 2 Pj and we proceed to the algorithm (3)–(6), if

where a ¼ 1 v KK0 , K ¼ k-complete compression module, v-porosity, K0 -the true

We calculate us ðM 0 Þ; M 0 62 S3Cs 2 Pj in all the layers j = 1, … n with the use of

Civil Engineering Department, Engineering Faculty, Gazi University,

Abstract. We present the assessment of a landslide that took place in the

© Springer Nature Switzerland AG 2019

2.2 Regional Geology and Tectonic Setup

Fig. 1. Views of the completed and failed ski jump complex

3.1 Assessment of Available Data

Fig. 4. Boreholes layout

Fig. 5. Slope arrangements of K125, K65, and K40 in runs